Labour supply and employment in the euro area countries - developments and challenges

Task force of the Monetary Policy Committee of the European System of Central Banks

Occas iOnal PaPer ser i e snO 87 / J une 2008

nO

87

Ju

ne

2008

Occ

asiO

nal

PaP

er s

erie

s

laBOur suPPlY anD

eMPlOYMenT in THe

eurO area cOunTries

DeVelOPMenTs

anD cHallenGes

OCCAS IONAL PAPER SER IESNO 87 / JUNE 2008

Task Force of the Monetary Policy Committee

of the European System of Central Banks

LABOUR SUPPLY AND EMPLOYMENT

IN THE EURO AREA COUNTRIES

DEVELOPMENTS AND CHALLENGES

This paper can be downloaded without charge from

http://www.ecb.europa.eu or from the Social Science Research Network

electronic library at http://ssrn.com/abstract_id=1084913.

In 2008 all ECB publications

feature a motif taken from the €10 banknote.

© European Central Bank, 2008

Address Kaiserstrasse 29

60311 Frankfurt am Main

Germany

Postal address Postfach 16 03 19

60066 Frankfurt am Main

Germany

Telephone +49 69 1344 0

Website http://www.ecb.europa.eu

Fax +49 69 1344 6000

All rights reserved. Any reproduction, publication or reprint in the form of a different publication, whether printed or produced electronically, in whole or in part, is permitted only with the explicit written authorisation of the ECB or the author(s).

The views expressed in this paper do not necessarily refl ect those of the European Central Bank.

ISSN 1607-1484 (print)

ISSN 1725-6534 (online)

3ECB

Occasional Paper No 87

June 2008

CONTENTS

EXECUTIVE SUMMARY AND CONCLUSIONS 6

1 INTRODUCTION 10

2 CONCEPTUAL ISSUES RELATING TO

LABOUR SUPPLY AND THE MACROECONOMY 12

3 MAIN TRENDS IN LABOUR SUPPLY 15

3.1 Population and the labour force 16

3.2 Participation rates 18

3.2.1 Participation rates by

individual characteristics in

the euro area 18

3.2.2 Participation rates in the

euro area countries 20

3.2.3 Compositional effects in

participation 21

3.2.4 Cohort effects in participation 22

3.3 Hours worked 27

3.3.1 Hours worked in the euro

area and in the United States 27

3.3.2 Hours worked in the euro

area by different worker groups 29

3.3.3 Hours worked per week in

the euro area countries 30

3.3.4 Compositional effects in

hours worked 31

3.4 Immigration 32

3.5 The supply of knowledge and skills 38

3.5.1 Educational attainment 38

3.5.2 The quality of educational

attainment 40

4 STRUCTURAL POLICIES AND THEIR

EFFECT ON LABOUR SUPPLY 45

4.1 Tax and benefi t systems 46

4.1.1 Unemployment benefi t

systems 46

4.1.2 Labour taxes 47

4.1.3 Pension systems 52

4.2 Parenthood and participation 56

4.2.1 Fertility and female

participation: what is the

relationship? 56

4.2.2 Availability and cost of

child care 57

4.2.3 Maternal and parental leave

opportunities 59

4.2.4 Part-time work

opportunities 59

4.3 Immigration Policies 61

4.3.1 Policies affecting

immigrants once in a country 63

4.4 Education systems and incentives 64

5 MATCHING LABOUR SUPPLY AND DEMAND 71

5.1 Returns to education 71

5.2 Relative unemployment rates of

different groups of workers 74

5.2.1 Institutions affecting the

matching of labour supply

with labour demand 74

5.2.2 Educational mismatch 76

5.2.3 Regional mismatch 79

5.2.4 Mismatch by age, gender

and nationality 79

5.3 The future demand for labour 83

ANNEX 1

MEASUREMENT ISSUES 85

ANNEX 2

DATA SOURCES AND DESCRIPTIONS 87

ANNEX 3

DETAILED TABLES AND CHARTS 91

ANNEX 4 108

REFERENCES 110

EUROPEAN CENTRAL BANK OCCASIONAL

PAPER SERIES SINCE 2007 121

LIST OF BOXES:

Box 1 Projections of labour supply in the

euro area and euro area countries 25

Box 2 Recent experiences with

immigration in euro area countries 35

Box 3 Human Capital and growth 40

Box 4 Labour quality growth in the

euro area and euro area countries 42

Box 5 Recent reforms of tax and benefi t

systems. Case study: Ireland and

Germany 50

CONTENTS

4ECB


June 2008

Box 6 Enhancing the participation rate of

older workers: the need for a

comprehensive strategy 53

Box 7 Reform of childcare and fl exible

working arrangements. Case study:

France and Finland. 60

Box 8 Incentives for fi rms and workers to

invest in human capital

accumulation and life-long

learning 68

Box 9 Returns to education in the

euro area countries 72

Box 10 The gender pay gap and

discrimination 82

Box 11 ILO defi nition of unemployment 85

Box 12 Cross-border commuting in the

euro area. Case study:

Luxembourg 108

JEL classifi cation: E5, J1, J2, J6

Keywords: Labour supply, employment, participation, hours worked, immigration, skill and

education, structural policies, labour demand, unemployment, euro area countries, labour markets,

taxes and benefi ts, childcare, pensions, training, human capital, labour quality, working time and

contracts, discrimination, mismatch, returns to education.

5ECB


June 2008

TASK FORCE OF THE

MONETARY POLICY

COMMITTEE OF THE

EUROPEAN SYSTEM

OF CENTRAL BANKS

TASK FORCE OF THE MONETARY POLICY COMMITTEE OF THE EUROPEAN SYSTEM OF CENTRAL BANKS

This report was drafted by an ad hoc Task Force of the Monetary Policy Committee of the European

System of Central Banks. The Task Force was chaired by Klaus Masuch. The coordination and

editing of the report was carried out by Jarkko Turunen (during the period March-August 2007)

and Melanie Ward-Warmedinger (during the period August 2007-June 2008). Clare Childs, Dung

Hoangkim and Stefanie Peuckmann provided editorial assistance.

The full list of members of the Task Force is as follows:

Klaus Masuch European Central Bank

Ramon Gómez-Salvador European Central Bank

Nadine Leiner-Killinger European Central Bank

Rolf Strauch European Central Bank

Jarkko Turunen European Central Bank

Melanie Ward-Warmedinger European Central Bank

Jan De Mulder Nationale Bank van België/Banque Nationale de Belgique

Harald Stahl Deutsche Bundesbank

Yvonne McCarthy Central Bank and Financial Services Authority of Ireland

Daphne Nicolitsas Bank of Greece

Aitor Lacuesta Banco de España

Mathilde Ravanel Banque de France

Piero Cipollone Banca d'Italia

Christelle Olsommer Banque centrale du Luxembourg

Alfred Stiglbauer Oesterreichische Nationalbank

Álvaro Novo Banco de Portugal

Klara Stovicek Banka Slovenije

Heidi Schauman Suomen Pankki

Other contributors:

Almut Balleer European Central Bank

Kieran McQuinn Central Bank and Financial Services Authority of Ireland

Pasqualino Montanaro Banca d'Italia

Alfonso Rosolia Banca d'Italia

Eliana Viviano Banca d'Italia

Cláudia Filipa Duarte Banco de Portugal

Matija Vodopivec Banka Slovenije

6ECB


June 2008

EXECUTIVE SUMMARY AND CONCLUSIONS

RATIONALE AND MAIN OBJECTIVE OF THE REPORT

The functioning of labour and product markets

affects the economic environment in which

monetary policy is conducted. For example,

structural policy measures which enhance

labour supply and employment growth increase

the pace at which an economy can grow

without higher infl ation. A greater fl exibility

of euro area labour markets and wages would

reduce adjustment costs and infl ation pressures

in the case of adverse supply shocks and

augment resilience of the economy, facilitating

the conduct of the stability-oriented monetary

policy of the ECB. In addition, an enhanced

fl exibility of wages and labour mobility is

needed to limit employment losses in the case

of adverse country specifi c shocks, thereby

facilitating the functioning of EMU.

Labour markets in the euro area, as in other

developed economies, face the triple challenge of

demographic change, technological progress and

globalisation. Demographic change (a decrease

in the size of the working age population and an

increase in the average age of the labour force)

reinforces the need to increase labour market

participation and employment in the euro

area. Increasing the share of people working

will help to support the euro area’s potential

output and per capita income, and reduce the

old-age dependency ratio. All this would help

to fi nance pension and health care systems

and to reduce the per capita fi nancing burden

for those who have to pay for these systems

via taxes and social security contributions.

Technological progress results in fi rms

searching for people with new types of skill and

knowledge. In order to support both high wages

and low unemployment, it is important that the

population of the euro area is well educated

and trained in the types of skills that fi rms

seek and that workers invest in enhancing and

developing their skills over the course of their

working lives. Effi cient schooling and education

systems, including vocational training, will play

an important role in enhancing investment in

education and equipping individuals (including

older persons) with both the skills demanded in

the labour market at any point in time and the

general competences that will allow them to

adapt fl exibly to new developments in labour

demand. In this context, training in skills sought

by traditional industries also remains important.

Finally, globalisation increases the ease with

which fi rms can either hire labour from abroad

(through immigration) and/or relocate their

production and services. It also increases the

competition faced by fi rms in the production

of goods and services. For workers in Europe,

this means that they have to remain competitive

in terms of the interplay between type and

level of skills, adaptability, productivity and

compensation packages.

Against this background and to successfully face

this triple challenge, conditions must be in place

that enhance the quantity and quality of labour

supply and effi ciently match the workforce with

fi rms’ demand for labour. This is necessary to

maximise individuals’ income and welfare and –

at the aggregate level – an economy’s potential

output, allowing an increase in the economy’s

rate of sustainable growth. Well-designed and

fl exible labour and product market institutions

are essential to this process. Moreover, structural

policy changes that enhance incentives for

schools, universities and fi rms to identify and

develop the “right” skills are needed. In addition,

the euro area should make the best use of global

labour supply through immigration, ensuring

that immigrants are effectively integrated into

its labour market and society.

The aim of this report, which has been prepared

by a Task Force of the Monetary Policy

Committee of the Eurosystem, is to describe

and analyse the main developments in labour

supply and its determinants in the euro area,

review the links between labour supply and

labour market institutions, assess how well

labour supply refl ects the demand for labour in

the euro area and identify the future challenges

for policy-makers. The data available for this

report generally cover the period from 1983 to

spring 2007. The cut-off date for the euro area

statistics included was 14 December 2007.

7ECB


June 2008

EXECUTIVE

SUMMARY AND

CONCLUS IONSMAIN FINDINGS ON DEVELOPMENTS IN

LABOUR SUPPLY:

Following a conceptual introduction to labour

supply from a macroeconomic perspective in

Chapters 1 and 2, Chapter 3 documents the key

developments in labour market participation,

employment and hours worked in the euro area

and EU Member States. Particular emphasis is

given to issues relating to the quality of labour

and human capital and immigration. Chapter 3

fi nds that:

i) over the period 1996-2007, overall labour

market developments appear to have been

quite favourable for the euro area as a whole.

Employment growth accelerated signifi cantly,

with the equivalent of 21.6 million new jobs

created. This contributed to a substantial

3.9 percentage point reduction in the

unemployment rate, which stood at 7.5% in

spring 2007. As a result, the positive gaps in

labour market participation and employment

rates between the United States and the euro area

became narrower and are now closer to those

prevailing in the early 1970s. In the euro area,

the total labour market participation rate rose by

5.6 percentage points over the period 1996-2007,

to nearly 71%, and the employment rate rose by

7.7 percentage points to over 65%. The highest

levels of participation in 2007 were registered

for men at 78% (employment at 73%), for prime-

aged individuals at 85% (79%), for citizens of the

new EU Member States (countries joining the

EU since 2004) at 77% (69%) and for the highly

educated at 88% (85%). Increases in participation

have compensated for a slight decline in the

working age population in recent years. However,

whilst much has been achieved, there is no room

for complacency. For example, labour supply

projections show that even if these positive

participation trends continue, they will soon no

longer be suffi cient to counteract the fall in the size

of the working age population. Furthermore, by

international standards, many euro area countries

continue to exhibit high unemployment rates and

low labour market participation;

ii) labour supply composition has changed over

time. Particularly women and the so-called

non-EU15 immigrants (immigrants from the

new EU Member States and non-EU countries)

have entered the labour market in increasing

numbers, and older workers have tended to

remain in the labour market for longer. The

participation of these groups increased by

9.0 percentage points, 7.4 percentage points

and 10.5 percentage points respectively

over the period 1996-2007. Changes in

educational levels, preferences and social

norms have, over time, played a role in

increasing female labour market participation

(through so-called cohort effects). Experiences

with immigration vary considerably by

country, but on the whole, immigrants have

contributed positively to labour supply and

employment, enhancing competition in labour

markets and helping to fi ll skills shortages.

Cross-border commuting within the euro

area has increased threefold in the last ten

years. Potential for further increases in labour

market participation exists, particularly among

younger workers (following completion of

their education), women and older workers, and

through enhancing immigrants’ integration;

iii) these developments have been accompanied

by an increase in labour quality and the

share of workers with higher education,

particularly those with tertiary level education

(by 6 percentage points between 1996 and 2007,

to a level of 20.7%). This implies an increase

of 16.7 million in the number of people with

tertiary level qualifi cations. The proportion of

low-skilled workers decreased in the euro area

over the period considered (by 9 percentage

points to a level of 37.7%). However, the level

of human capital in some euro area countries

lagged behind that in other advanced countries,

and labour quality growth seems to have slowed

towards the end of the 1990s, highlighting

the need for further increases in educational

attainment and on-the-job training; and

iv) as the number of workers has increased,

the average hours worked per week has

declined in the euro area (by a total of

1.2 hours over the period 1996-2007). This

refl ects changes in working time regulations

8ECB


June 2008

and especially an increase in part-time jobs,

which has helped women, in particular, to join

the labour market in greater numbers. The ratio

of part-time to total employment increased

by around 5 percentage points over the

same period.

MAIN FINDINGS ON THE STRUCTURAL POLICIES

DETERMINING PARTICIPATION AND EMPLOYMENT

AND POLICY CONCLUSIONS:

Chapter 4 reviews how structural policies affect

labour supply and may have contributed to the

developments observed. Chapter 5 presents

evidence on how well the characteristics of

labour supply have refl ected those of the

demand for labour over the last two decades and

discusses how to respond to the main challenges.

The welcome developments in participation and

employment rates over the last decade partly

refl ect increased labour market fl exibility and

reform progress. However, this progress has been

quite uneven across countries. Looking forward,

a priority for economic policies is to ensure high

potential output, greater fl exibility and resilience

of the euro area countries to shocks. While there

is often no single solution for all, it is important

that countries learn from each other and from

best practices to develop new and better labour

market institutions and policies which will help

to achieve desired results. Identifying which

economies have performed best with regard to

which specifi c policy areas should help with

this. Furthermore, it is important to consider

reform packages as a whole when embarking

on reform processes, in order to predict trade-

offs in a timely manner and ensure that policies

are complementary within the framework of a

comprehensive reform strategy. The key fi ndings

and policy conclusions of the report are as

follows:

i) The analysis undertaken in Chapter 4 shows that

structural policies affect, inter alia, individuals’

labour market decisions and household income.

There is a need to further optimise policies in the euro area and to increase the labour market participation and employment of all groups, especially females and non-prime-aged

workers. Reducing high marginal tax rates and

tax wedges, high unemployment benefi t levels

and long durations, weak work availability

requirements and early retirement schemes

would stimulate labour supply and income.

Tax and benefi t systems should not discourage

older workers from voluntarily staying longer

in the labour market (for example, fi nancial

disincentives for extending participation beyond

standard retirement ages should be removed).

Reducing taxes and social security contributions

on labour, especially if funded by expenditure

constraint and enhanced effi ciency of public

fi nances, would enhance employment and net

wages, as well as provide additional incentives

for individuals to move from unemployment to

work or to invest in human capital. Restrictions

on working arrangements and labour contracts

should be reduced to allow for more fl exible

working hours (both higher and lower

than standard contracts) and more fl exible

employment protection. Furthermore, so called

“work/family reconciliation” policies that

facilitate the combination of work and a family

(such as the provision of affordable childcare,

parental leave and part-time work opportunities)

should be well-designed to support and

encourage labour market participation and

ensure equal opportunities. Evidence suggests

that the success of reform measures hinges on

the use of comprehensive reform strategies that

take into account the factors that may infl uence

individuals’ decisions to work and the ease with

which they fi nd a job.

ii) Chapter 5 shows that labour market regulations and institutions need to be more fl exible to facilitate the matching of labour supply with labour demand (across skills,

worker groups, sectors and regions). Labour

market institutions should be geared towards

lowering adjustment costs (e.g. by reducing

employment protection) and barriers to cross-

occupation and geographical labour mobility

(e.g. through common educational standards

and easier pension transfers) and should offer

appropriate incentive-compatible fi nancial

support for those temporarily unemployed and

job search assistance. Flexible wage bargaining

9ECB


June 2008

EXECUTIVE

SUMMARY AND

CONCLUS IONSis important for allowing wages to refl ect local

labour market conditions (such as regional and

skill-specifi c unemployment rates, regional and

sectoral productivity growth and workers’ skills).

Wages that are not suffi ciently differentiated –

for example, by skill or region – exacerbate

the mismatch between labour supply and

labour demand (also by not providing the

incentives for capital to shift to areas with

high unemployment), thus increasing the

unemployment rates of certain skill groups and

regions. Institutions have an important role

to play in matching labour supply with labour

demand. Public employment services need to

be more effi cient. Institutional arrangements

that hinder the employment of low-skilled

workers (such as excessive minimum wages)

should be avoided. Contractual arrangements

should give individual workers greater freedom

to agree on contract details (e.g. working hours

and options to invest in additional training).

Moreover tight product market regulation,

and more generally measures that hinder or

restrict competition, are an impediment to job

creation. Policies that increase competition in

goods and service markets, such as those that

reduce the administrative burden on fi rms start-

ups and remove statutory barriers to entry in

certain sectors, would help support employment

creation.

iii) Chapters 3, 4 and 5 highlight the need to increase skills and knowledge (and thus the quality of labour supply) and the transferability of skills. Employment normally enhances skills

because workers learn by doing. Therefore,

bringing unemployed or inactive people into

jobs will, over time, enhance individuals’ labour

productivity and thus real wages. An effi cient

framework for training and counselling the

(long-term) unemployed should help them

to remain employable. The large differences

in tertiary education funding between the

United States and the euro area should be

addressed also by enhancing conditions for

private funding. In addition, the labour market

(including fi rms) should play a stronger role in

signalling to education systems and workers

which skills are expected to be in short supply.

Good quality education should be ensured, in

particular, by enhancing the relevant incentives

and recognition for young people, workers and

fi rms to invest in education and training. The

effi ciency and service orientation of education

institutions should be improved, e.g. by

enhancing some elements of competition and

external quality controls.

iv) Evidence presented in Chapters 3, 4 and

5 also shows that the euro area should make better use of skills from outside the euro area.

However, immigration is not a substitute for

economic reform geared towards removing

barriers to high labour market participation.

Rather, it provides a means through which

the euro area can tap into the global supply

of labour resources and fi ll domestic skill

shortages. From an economic perspective,

the benefi ts derived from immigration depend

on the characteristics of migrants entering

the euro area and the ease with which they

integrate into work and society. Immigration

policy should be closely aligned with the

skills needed by the labour market, ensuring

appropriate migration fl ows that have the

potential to fi ll gaps in labour supply and ease

adjustments over time (particularly in view

of population ageing). Selective migration

policies which limit labour mobility within

the EU should be avoided and replaced

by measures that support labour market

mobility (such as the increased portability

of pension rights). Policies that ensure the

successful integration of immigrants into

the active workforce and society as a whole

(e.g. through incentives to broaden language

skills) are crucial. Successful integration

is also important because some migratory

fl ows are not steered in line with immediate

skill needs (such as family reunifi cation and

asylum seekers). In some countries, reforms

may be necessary before larger numbers of

immigrants can be expected to be integrated

successfully into labour markets.

10ECB


June 2008

1 INTRODUCTION 1

This report aims to analyse the main

developments in labour supply and its

determinants in the euro area, describe the level

and structure of employment, 2 review some of

the links between labour supply, labour market

institutions and economic performance, assess

how well labour supply refl ects the demand for

labour in the euro area, and identify challenges

relating to these topics for policy-makers.

These issues are very important for economic

development and welfare more generally, both

at an individual and aggregate level. They are

also important from a central bank perspective,

since labour supply affects the environment in

which monetary policy is conducted. The precise

impact of changes in labour supply on potential

and actual output, as well as on the natural and

actual unemployment rate, is affected by the

design and fl exibility of labour and product

market institutions and by the adjustment of

labour costs to these developments. In this

report labour supply is understood in a broad

sense to cover the size and composition of the

labour force (including the self-employed) by

age, gender, educational attainment level and

nationality (as an indicator of immigrant status)

and is thus analysed in terms of both quality and

quantity.

The main developments in euro area labour

markets, from a long-term perspective, are a

signifi cant decline in employment rates and an

increase in unemployment rates, which started

in the 1970s. Over the most recent decade

developments reversed, leading to rising labour

force participation and employment rates, as

well as declining unemployment rates. Policy

initiatives, such as the European Employment

Strategy and the Lisbon Agenda for Growth

and Jobs, and a favourable macroeconomic

environment, have supported these recent

improvements. Many euro area countries have

made some progress with labour market reforms,

such as improving work incentives, and by

reducing product market regulations, which also

affect labour market performance. However, this

progress has been quite uneven across countries

and further reform of labour and product markets

is needed. By international standards, most euro

area countries still have high unemployment

rates and low labour market participation.

Such levels cannot be explained by factors like

cyclical development, suggesting that there

are still structural and institutional barriers to

labour supply and employment within the euro

area. In this respect, it is possible to identify a

number of specifi c labour supply characteristics

in the euro area (and, more widely, Europe)

that warrant further consideration. These

include low (but increasing) female labour

supply rates, a relatively low youth labour force

participation rate but an increasingly educated

workforce, relatively high early retirement rates,

recent increases in immigration, high youth

unemployment rates and the existence of labour

market mismatches across certain groups of the

labour force, regions and skills.

This report is organised as follows: Chapter 2

briefl y discusses labour supply in the

macroeconomy from a conceptual point of view.

It details how developments in labour supply are

relevant to economic developments, welfare and

monetary policy, outlining the possible impact

of changes in both the quantity and quality of

labour supply on wages and output over the

short to medium term. Institutions’ conceptual

role in shaping this impact is also discussed.

Chapter 3 presents empirical evidence on the

main trends in the quantity and quality of labour

supply and employment in the euro area and

euro area countries over the last two decades.

It considers how the age, gender, educational

attainment and nationality profi le of labour

supply in the euro area has changed over time

by assessing the participation and employment

of these sub-groups. Particular emphasis is

given to developments in immigration and the

supply of human capital. Chapter 4 reviews

how structural policies affecting labour supply –

namely tax and benefi t systems, work-family

Prepared by J. Turunen and M. Ward-Warmedinger.1

Providing a comprehensive analysis of the determinants of 2

employment would require a deeper analysis also of labour

demand issues and falls outside the scope of this report.

11ECB


June 2008

I INTRODUCTION

balance policies (including the provision of

childcare and part-time work opportunities),

immigration and educational policy – may have

helped shape the labour supply developments

of the sub-groups considered in the previous

chapter. It presents qualitative assessments of

key reforms and structures and their impact

on the quantity and quality of labour supply.

Looking forward, policies affecting labour

supply will need to accommodate a number of

challenges. Chapter 5 presents evidence on how

well labour supply has refl ected the demand for

labour over the last two decades by analysing

the returns to education and the level and

change in unemployment rates by skill, region,

age, gender and nationality. It presents a brief

overview of how labour and product market

institutions affect the matching of labour supply

with labour demand. Finally, it discusses the

implications of globalisation, demographic and

technological change for the future composition

of labour supply.

12ECB


June 2008

2 CONCEPTUAL ISSUES RELATING TO LABOUR

SUPPLY AND THE MACROECONOMY 3

Labour supply developments, in terms of size,

quality and composition, are a major determinant

of an economy’s potential output, affecting an

economy’s rate of sustainable growth. Increasing

the share of people and skills in work will also

help support per capita income and reduce the

old age dependency ratio, which would help

reduce the fi scal burden related to population

ageing. Furthermore, the cyclical sensitivity of

labour supply can affect labour market tightness

and thereby infl uence the outlook for wages

and prices and infl ation dynamics over the

business cycle frequency. The labour supply’s

precise impact on potential and actual output,

and on the natural and actual unemployment

rate, is affected by the design and fl exibility of

labour and product market institutions and the

response of labour cost developments. Sub-

optimal structural and fi scal policy measures

may undermine productivity and increase

structural unemployment, with consequences

for monetary policy. Such factors also affect

individuals’ decisions to supply labour and

the types of labour supplied (discussed further

in Chapter 4). Changes in the composition of

labour supply 4 relative to demand can have

important consequences for the unemployment

rates of particular groups of workers (and

thus total unemployment rates), especially if

labour markets or wages are not suffi ciently

fl exible (discussed further in Chapter 5). These

factors infl uence labour supply developments’

impact on actual and expected wage and price

pressures, with possible implications for the

conduct of monetary policy. This chapter briefl y

reviews how developments in labour supply

affect the macroeconomy and their relevance

for monetary policy.

Labour supply is a key contributor to

economic growth. Both an increase in labour

input as measured by total hours worked

(employment times hours worked per worker,

sometimes referred to as labour utilisation)

and improvements in human capital have the

potential to contribute positively to real GDP,

income growth and welfare. Labour utilisation

is largely determined by developments in

population growth (itself a function of fertility

and mortality rates and immigration, discussed

further in Box 1) and the likelihood that those

in the working age population participate in the

labour market (dependent on job opportunities

and incentives/disincentives to enter the labour

market, discussed further in Chapters 4 and 5).

Furthermore, hours worked from a long-run

perspective – that is total hours of work over an

individual’s lifetime – affect the level of output.

The potential for the supply of human capital

(encompassing factors such as the quantity

and quality of formal education, labour market

experience and on-the-job training, as well as

a broader set of competencies, e.g. cognitive

abilities) to contribute to growth appears

substantial as refl ected in the prominent role

of human capital in modern growth theory.

In particular, endogenous growth models

suggest that improvements in human capital

can generate technological progress and

thus productivity growth in the long term.5

Human capital contributes to measured total

factor productivity growth through changes

in the skill composition of the workforce and

possible interactions between human capital and

technology adoption.6

Changes in labour supply at the business cycle

frequency can affect labour market tightness.

Besides the magnitude and the persistence of a

positive or negative labour supply shock, the

extent and propagation of its impact on

macroeconomic variables depends on its

interaction with the economy’s institutional

framework. Three main categories of rigidities

infl uence the transmission of a labour supply

shock (and on the short-run unemployment-

infl ation trade-off), namely real wage rigidity,

the rigidity of contracts (e.g. contract duration

Prepared by K. Stovicek, J. Turunen and M. Ward-3

Warmedinger.

See Section 3.2.3 and section 3.2.4 for a discussion of 4

composition effects in labour market participation and their

effect on aggregate levels and trends in labour supply.

Barro and Sala-i-Martin (2004).5

See Gomez-Salvador et al. (2006) for a more detailed description 6

and further references.

13ECB


June 2008

2 CONCEPTUAL ISSUES

RELATING TO LABOUR

SUPPLY AND THE

MACROECONOMY

and contract design fl exibility) and rigidities in

the legislative and economic framework

(e.g. employment protection).7 The interaction

of a positive labour supply shock with rigid

labour market institutions may shift adjustment

over the business cycle from prices (wages) to

quantities, thus increasing unemployment in

particular in the short to medium run.

Furthermore, lower incentives to take up a job

(e.g. stemming from disincentives created by

the unemployment insurance system), as well as

regulations restricting labour demand, such as

high minimum wages, may contribute further to

a larger accommodation of increased labour

supply via unemployment. From this perspective,

labour and product market reforms (that, for

example, reduce red tape, enhance real wage

fl exibility, lower employment protection,

increase incentives from unemployment

insurance systems etc. – see Chapter 4) may

contribute to an economy’s shock absorption

capacity, by either dampening the unemployment

effect of the labour supply shock, or speeding

the economy’s return to a high employment

equilibrium by lowering the persistence of

employment and output fl uctuations.8 Some of

these structural reforms may also help lower the

natural rate of unemployment.

The measured participation rate tends to be pro-

cyclical, since persons on the edge of labour

market attachment (who have acquired little

work experience or career-specifi c education 9)

and who are more likely to move in and out of

the labour market react to changes in labour

market conditions and job availability (discussed

further in Chapter 4). The degree to which labour

supply adjusts through participation is also

affected by market rigidities, job availability

and matching frictions. Movements in and out

of the labour force may complicate the task

of measuring labour market participation and

unemployment with precision, since changes in

the measured participation rate may not refl ect

a change in individuals’ preferences (since, for

example, labour market rigidities may mean

that those wishing to work are discouraged

from participating in the labour market).

Issues related to the diffi culty of measuring

labour market participation and unemployment

(arising from, inter alia, actual moves in and

out of the labour market or from work in the

informal economy) are discussed further in

Annex 1 and Box 11. Average hours worked

(per person employed) are also procyclical

since they provide an adjustment mechanism

for employment. This form of adjustment may

be particularly relevant for fi rms expecting a

change in product demand to be short-lived and

operating in the labour market with substantial

adjustment costs to employment (in terms of

persons employed). The shift from full-time to

part-time employment, and vice versa, is another

driver of changes in hours worked.

Changes in the composition of labour supply

relative to demand also have important

consequences for the wage structure and

unemployment rates of sub-groups of workers.

Assuming unchanged demand for each labour

type and fl exible wages, an increase (decrease)

in the relative supply of a specifi c type of

workers results in a lower (higher) wage for

that type, relative to other types. If, however,

relative wages are not completely fl exible, the

change in relative supply results in changes in

relative unemployment.

Finally, changes in labour supply naturally have

implications for monetary policy, since they

may affect the actual and natural rates of

unemployment and infl ation dynamics. The

institutional framework is also important in this

context, since it can affect both the transmission

of a labour supply shock to infl ation dynamics

(in particular through its affect on wage

developments and their pass-through to prices)

and the transmission of monetary policy to

infl ation itself (through its effect on adjustment

in the labour market). Recently, the development

See Layard et al (2005).7

As stated in Duval, Elmeskov and Vogel (2007), there is no 8

simple link between rigidities in labour and product markets and

resilience, since institutions that dampen the initial impact of a

shock may also increase its persistence and vice versa. Therefore

the net effect of structural policies remains an empirical issue.

See e.g. Aaronson, Fallick, Figura et al. (2006), Bradbury 9

(2005), Elmeskov and Pichelman (1993), Clark et al (1979).

14ECB


June 2008

of micro-founded structural models to

incorporate various labour market features

shows that labour market rigidities, notably

wage rigidity, increase infl ation persistence,

thereby changing the short-run trade-off between

infl ation and unemployment.10 In this framework,

labour market rigidities may affect an optimal

monetary policy aiming to reduce the welfare

costs of macroeconomic fl uctuations.11 In a more

fl exible labour market, wages could be expected

to more closely refl ect workers’ marginal

productivity, reducing the impact of a labour

supply shock on short-term infl ation.

Within the literature to date, labour supply shock transmission is 10

strongly dependent on a model’s characteristics.

Blanchard and Gali (2007) explore the transmission of a 11

productivity shock in the framework of real wage rigidity. They

fi nd that to the extent this transmission has implications in the

medium term, optimal monetary policy may take into account

both infl ation and the output variability. Christoffel, Kuester

and Linzert (2006) fi nd that in the presence of labour market

frictions and wage rigidity, an optimal monetary policy might

consider both wage and price infl ation in their monetary policy

reaction function.

15ECB


June 2008

3 MAIN TRENDS IN

LABOUR SUPPLY3 MAIN TRENDS IN LABOUR SUPPLY12

This chapter presents empirical evidence on the

main trends in the quantity and quality of labour

supply in the euro area from the early 1980s

onwards using data from Eurostat’s Labour Force

Survey (EU-LFS).13 Labour supply developments

are discussed in terms of changes in population,

labour market participation and hours worked by

four main individual characteristics: age, gender,

education and nationality. The chapter also briefl y

covers measurement issues relating to the main

indicators of labour supply, with emphasis on the

measurement of human capital and the quality

of labour supply. Finally, this chapter describes

recent developments in immigration with a focus

on the euro area countries where migration has

been particularly important for labour supply.

When using data on employment, unemployment

and the labour force, it is important to take note

of some measurement issues that can potentially

lead to some mis-measurement of the true levels

of these variables. For instance, a number of

individuals working in the shadow economy and

in household production are not registered as

employed. Moreover there are individuals who do

not work but are available to work, and may not

be captured by the defi nition of unemployment

used to collect labour market statistics in the

EU-LFS, resulting in an underestimation of the

actual labour supply (see Annex 1 for a more

detailed discussion). For example, since non-

participation and the number of hours worked

are not just the result of individual preferences,

but also of market rigidities which undermine

labour demand, “true” unemployment in the

euro area may be higher than measured.14 Taking

these issues into account in a systematic way is

beyond the scope of this study. However, given

these potential measurement problems regarding

the distinction between non-market activities,

inactivity and unemployment, and the overall

importance of employment, this report also

presents information on employment rates.

Main fi ndings of this chapter include an

increase in labour market participation of

5.6 percentage points in the euro area over the

period 1996 to 2007, with the highest levels of

participation in 2007 registered for men (78%),

prime-aged individuals (85%), non-nationals

from the 12 new Member States (77%) and the

highly educated (88%). In particular, women

and non-EU15 immigrants have entered the

labour market in increasing numbers, and older

workers have tended to stay in the labour market

longer. The labour market participation of these

groups increased by 9.0 percentage points,

7.4 percentage points and 10.5 percentage

points respectively (0.9 percentage point,

0.6 percentage point and 0.9 percentage point

on average per year). Cohort effects linked

to changes in educational levels, preferences

and social norms over time played a role in

increasing female labour market participation.

Over this period, developments in total hours

worked in the economy, as an alternative

measure of labour input, show an upward trend

similar to that observed for employment. At the

same time, average weekly hours of work per

employed person in the euro area have declined

by 1.2 hours per week over the period 1996 to

2007, largely due to the increase in part-time

jobs, which increased by around 5 percentage

points as a share of total employment. These

developments have been accompanied by an

increase in the share of the population with

higher education, in particular those with

tertiary level education (by 6 percentage

points since 1996, to 20.7% in 2007). The

proportion of the population with a low level

of education has fallen (by 9 percentage points

over the same period, to 37.7% in 2007). These

positive developments have led to an increase

in participation rates and have compensated

for the negative impact on labour supply of the

Prepared by R. Gomez-Salvador.12

See Annex 2 for a description of this dataset. The need to 13

achieve international comparability means that the LFS dataset

uses standardised and widely accepted defi nitions of e.g.

employment and unemployment, as adopted by ILO. These

constitute the basis of the Eurostat LFS. It should be noted that

these defi nitions differ from those adopted by countries in their

national defi nitions of labour market status, where international

comparability is not necessary.

Suggesting, for example, underemployment-that is, the lower 14

hours worked per year-in the euro area may not be voluntary

(or may not be a matter of choice). See Leiner-Killinger,

Madaschi and Ward-Warmedinger (2005) for a discussion of

institutional arrangements reducing hours of work.

16ECB


June 2008

slowdown in population growth rates. However,

projections of future labour supply show that

under current policies, the labour force will start

declining soon due to a fall in the size of the

working age population.

3.1 POPULATION AND THE LABOUR FORCE 15

In 2007, the euro area labour force (the

employed plus unemployed) included over

148 million people, out of a total population

of 318 million and a working age population

(ages 15 to 64) of around 209 million. This

implies a participation rate (labour force divided

by working age population) of close to 71%.

Of the labour force, the number of employed

persons reached around 137 million in 2007,

leading to an employment rate (employment

divided by working age population) of 65.5%,

and the number of unemployed persons

was around 11 million, translating into an

unemployment rate (unemployment divided

by labour force) of 7.5%. Both euro area

participation and employment rates were below

those recorded in the United States (75.3%

and 71.8% respectively in the United States in

2007), while the euro area unemployment rate

was higher (4.7% in the United States) – see

last column of Table 1.

Table 1 also summarises recent and past

developments in population and labour market

indicators. Developments are presented in two

ways: trend developments, to control (to the

extent possible) for cyclical effects; and recent

developments, which divide the past decade

into two fi ve-year periods (of relatively higher

and lower economic growth – see average real

Prepared by R. Gomez-Salvador.15

Table 1 Population, working age population, participation, labour force, employment and unemployment in the euro area and the United States

(average year-on-year growth rates (%), unless otherwise indicated)

Trend developments Recent developments Level 1)

1984-1995 1996-2007 1996-2001 2002-2007 1983 2007

Euro area Population 0.3 0.4 0.3 0.6 291.9 318.8

Working age population 0.5 0.4 0.3 0.5 188.5 209.3

Participation rate 2) 0.2 0.5 0.4 0.6 63.3% 70.8%

Labour force 0.8 1.1 0.9 1.3 119.3 148.3

Population effect 3) 0.5 0.4 0.3 0.5

Participation rate effect 3) 0.3 0.7 0.6 0.8

Employment 0.6 1.6 1.6 1.4 107.2 137.1

Total hours n.a. 1.3 1.1 1.3 n.a. 222,400

Unemployment 1.9 -2.1 -4.2 0.1 12.1 11.1

Employment rate 2) 0.1 0.6 0.7 0.6 56.9% 65.5%

Unemployment rate 2) 0.1 -0.3 -0.6 -0.1 10.1% 7.5%

Real GDP 2.8 2.5 2.8 1.9

US Population 1.1 1.1 1.2 1.0 234.3 302.6

Working age population 1.1 1.4 1.4 1.3 148.3 195.6

Participation rate 2) 0.3 -0.1 0.0 -0.3 73.2% 75.3%

Labour force 1.5 1.2 1.4 1.0 108.5 147.3

Population effect 3) 1.1 1.4 1.4 1.3

Participation rate effect 3) 0.4 -0.2 0.0 -0.3

Employment 2.0 1.3 1.6 1.0 97.9 140.4

Total hours n.a. 1.2 1.3 1.0 n.a. 267,427

Unemployment -2.6 -0.4 -1.3 0.5 10.6 6.9

Employment rate 2) 0.5 -0.1 0.1 -0.2 66.0% 71.8%

Unemployment rate 2) -0.3 -0.1 -0.1 0.0 9.8% 4.7%

Real GDP 4.0 3.7 3.9 2.8

Sources: Eurostat, BLS and ECB calculations. Note: Euro area data refer to the second quarter of each year, while US data are annual averages. 1) In millions, unless otherwise indicated. 2) Average year-on-year changes (percentage point). 3) Contributions to average year-on-year growth rates (percentage points).

17ECB


June 2008

3 MAIN TRENDS IN

LABOUR SUPPLY

GDP in Table 1). In the past decade, overall

labour market developments have been quite

favourable for the euro area as a whole. As

a result, the positive gap in participation

and employment rates between the United

States and the euro area has narrowed and

is now closer to those prevailing in the early

1970s (see Chart 1). Positive developments

refl ect a particularly strong increase in female

participation, while male participation rates

actually fell below levels prevailing in 1983.

These developments took place in a context

of broadly stable growth of the euro area’s

working age population.

Labour force developments can be

decomposed into two effects. First, the

population effect, i.e. changes in the working

age population for given participation rates

and, second, the participation rate effect,

i.e. changes in the participation rate for a

given working age population. Comparing

average annual growth rates in the period

1996 to 2007 with those in the period 1984

to 1995, the increase in the participation

rate effect (0.4 percentage point – from

0.3 percentage point to 0.7 percentage point)

more than compensated for the slight trend

decline in the growth rate of the working

age population (the population effect

-0.1 percentage point – from 0.5 percentage

point to 0.4 percentage point), allowing the

growth rate of the labour force to increase

(from 0.8% to 1.1%).

Looking at the same developments within the

past decade shows two interesting results. First,

the participation rate contribution has increased

over time, something that contrasts with the

expected pro-cyclicality of participation rates,

Chart 1 Overall participation and employment rates for the euro area and the United States

(percentages)

Euro area

United States

Participation rate Employment rate

60

65

70

75

80

60

65

70

75

80

1970 1976 1982 1988 1994 2000 200655

60

65

70

75

55

60

65

70

75

1970 1976 1982 1988 1994 2000 2006

Participation rate males Participation rate females

60

70

80

90

60

70

80

90

1983 1986 1989 1992 1995 1998 2001 2004 200645

55

65

75

45

55

65

75

1983 1986 1989 1992 1995 1998 2001 2004 2006

Sources: European Commission, Eurostat, BLS and ECB calculations. Notes: 15 to 64 year olds. See Chart 11 in Appendix 3 for employment rates of males and females.

18ECB


June 2008

as periods of relatively low economic growth

normally tend to discourage workers from

participating in the labour force (as discussed

in Chapter 2).16 This suggests that labour

market developments have recently been related

to factors independent of the cycle, such as

changes in the composition of the labour force

and cohort effects. Second, the contribution

from population growth also increased, in

contrast with the trend decline in population

growth rates observed since the early 1980s.

3.2 PARTICIPATION RATES 17

3.2.1 PARTICIPATION RATES BY INDIVIDUAL

CHARACTERISTICS IN THE EURO AREA

The participation rate, and its evolution over

time, is not the same for all groups inside the

working age population. Characteristics such as

gender, age, qualifi cation level and origin have

a strong impact on the observed rate of labour

market participation. The EU-LFS provides

detailed information on the socioeconomic

characteristics of the working, unemployed and

inactive populations in the euro area. Data on

gender and age are available, and the survey

provides information on level of education,

by distinguishing between low (completion of

lower secondary education or less), medium

(completion of up to a diploma of upper

secondary education) and high (holding a

diploma of tertiary education) qualifi cation

level.18 Information on work experience,

training or on-the-job qualifi cations is not

available. Furthermore, information on

country of origin is limited to the nationality

of the survey respondent 19 and distinguishes

The cyclical behaviour of participation in the euro area is 16

documented in Genre and Gomez-Salvador (2002).

Prepared by J. De Mulder.17

The precise defi nition of the educational levels is provided in 18

Annex 2.

Nationality and origin do not necessarily provide the same 19

information, as immigrants or their descendants can have

obtained the nationality of the considered country.

Table 2 Euro area participation rates according to different subdivisions

(employment rates in brackets)

Average annual change (percentage points) Level (%) Trend developments Recent developments

1984-1995 1996-2007 1996-2001 2002-2007 2007

Total 0.2 (0.1) 0.5 (0.6) 0.4 (0.7) 0.6 (0.6) 70.8 (65.5)

Excluding the effect of changes in the population composition 1) n.a. (n.a.) 0.3 (n.a.) 0.1 (n.a.) 0.4 (n.a.)

According to gender

Males -0.3 (-0.4) 0.2 (0.3) 0.1 (0.4) 0.2 (0.2) 78.4 (73.2)

Females 0.6 (0.5) 0.7 (0.9) 0.6 (0.9) 0.9 (0.9) 63.2 (57.8)

According to age

15-24 years old -0.7 (-0.5) 0.0 (0.3) 0.0 (0.6) 0.0 (0.0) 44.0 (37.3)

25-54 years old 0.5 (0.2) 0.4 (0.6) 0.3 (0.7) 0.5 (0.5) 84.6 (79.1)

55-64 years old -0.3 (-0.4) 0.9 (0.9) 0.2 (0.3) 1.5 (1.5) 46.4 (43.4)

According to education level 2)

Low n.a. (n.a.) 0.4 (0.6) 0.3 (0.7) 0.5 (0.5) 63.5 (57.6)

Medium n.a. (n.a.) 0.2 (0.4) 0.1 (0.3) 0.4 (0.4) 80.4 (75.3)

High n.a. (n.a.) 0.0 (0.2) -0.1 (0.3) 0.2 (0.2) 88.3 (84.7)

According to nationality 3)

Nationals n.a. (n.a.) 0.3 (0.5) 0.4 (0.8) 0.2 (0.3) 70.9 (65.9)

Other EU15-citizens n.a. (n.a.) 0.2 (0.3) 0.0 (0.5) 0.4 (0.2) 73.7 (67.6)

Non EU15-citizens n.a. (n.a.) 0.6 (0.8) 0.2 (0.5) 1.1 (1.1) 69.6 (59.3)

of 12 new EU member states n.a. (n.a.) n.a. (n.a.) n.a. (n.a.) n.a. (n.a.) 77.1 (68.7)

of non EU27-countries n.a. (n.a.) n.a. (n.a.) n.a. (n.a.) n.a. (n.a.) 68.3 (57.7)

Sources: EU-LFS (spring data), ECB and NBB calculations. Note: 15 to 64 years old, except for the subdivision according to education level (25-64 years old). EU15 refers to those countries that were EU Members prior to 2004. The 12 new EU Member States include the countries joining the EU since 2004. 1) Calculated by weighting the participation rates of 18 subgroups of the population of working age (subdivided according to gender, age and education level) with the structure of the population of working age in 1995. 2) EU-LFS data concerning education level only available from 1992 onwards. 3) EU-LFS data concerning nationality only available from 1995 onwards.

19ECB


June 2008

3 MAIN TRENDS IN

LABOUR SUPPLYbetween nationals (persons having the

nationality of the considered country), other

EU citizens and citizens of other (non-

EU) countries. Although level of education

and nationality are not perfect measures

of qualifi cation and country of origin,

they provide the best available proxies for

constructing a long time series to investigate

developments in labour supply according to

these characteristics.

Consideration of the participation rates of these

different groups shows that, on average, female,

low-skilled, young and older persons and non-

EU citizens participate less in the labour market

than other groups (see Table 2). However, with

the exception of the young, the increase in the

participation rate of these groups has accelerated

over the last decade, with the result that they have

at least partially caught-up to the participation

level of other groups. Over the period 1996 to

2007, the average increase in participation has

been largest for females (0.7 percentage point),

older workers (0.9 percentage point) and non-EU

citizens (0.6 percentage point). In contrast, the

participation rate of 15-24 year olds stabilised

during this period.

Participation and employment rates are

still highest for males, prime-aged workers

(25-54 year olds), the highly educated and

EU citizens. Participation is substantially

higher for males than for females in all age

groups (see Table 27, Annex 3). Participation

for both genders is highest between the ages

25 to 49, but while some 90% or more of males

in this age group participated in the labour

market in 2007, this was only the case for

about 77% of females. Female participation

is strongly affected by family status. Until

the age of 49, female participation rates are

clearly lower when a woman has a partner and

when there are dependent children. Similar

differences between genders are also found for

employment rates.

The lower participation rate for younger

persons is often linked to their pursuit of

education. If this results in individuals

obtaining a diploma of upper secondary or, in

particular, tertiary education, the immediate

downward effect on participation rates of

studying longer is compensated afterwards

by higher labour market participation (and

employment) as the education level rises.

Rates of labour market participation, and in

particular employment, remain highest for

the highly educated, but the participation

rate gap related to education is gradually

getting smaller. This catch-up is attributable

to females. Their participation increased for

all three education levels, but the increase

was stronger for those with the lowest level

of education.

Nevertheless, the positive impact of higher

education on labour market participation is still

much stronger for females. In 2007, moving

from a low to medium education level increased

the female participation rate by 24 percentage

points, and from medium to highly skilled by

another 9 percentage points. For males, the

increases were 9 and 5 percentage points

respectively. The remaining gap between female

and male participation is therefore mainly due

to the low skilled. In 2007, 78% of 25-64 year

old low-skilled males participated in the labour

market, 20 while only 50% of their female

counterparts did the same.

Turning to the breakdown by nationality, in

2007, differences in labour market participation

across nationality groups within countries were

relatively small; the highest participation (and

employment) rate is found for citizens of the

12 new member states of the EU. The apparently

almost equal participation of nationals and non-

nationals nevertheless hides two important

facts. First, compared with the corresponding

fi gures for EU citizens, the labour market

participation of non-EU nationals is especially

low for women and middle-aged persons.

Perhaps surprisingly, this is also the case for

highly skilled non-EU citizens, whose

participation rate is comparable to that of

The participation behaviour of older people is treated in Box 6.20

20ECB


June 2008

medium-skilled EU-citizens, and particularly

for highly skilled females.21 In relative terms,

this group seems to experience substantially

more problems entering the labour market than

do less skilled immigrants.22 Second, although

non-EU-citizens do not appear to participate in

the labour market much less than nationals on

average, they are less likely to be employed (in

2007 by 8.2 percentage points), and are thus

signifi cantly more often unemployed (discussed

further in Chapter 5).

3.2.2 PARTICIPATION RATES IN THE EURO AREA

COUNTRIES

Labour market participation rates also vary across

euro area countries, ranging in 2007 from 78.5%

in the Netherlands to 62.5% in Italy (see Table 3).

At about 76% to 77%, participation rates were

also relatively high in Germany and Finland,

while Greece, Luxembourg and Belgium were

among countries with the lowest participation

rates at 67% or less. In comparison, employment

rates ranged from 76% of the working age

population in the Netherlands to 59% in Italy. By

means of comparison, in Denmark, Sweden and

the UK, levels of participation and employment

rates are still much higher than in the euro area

and only comparable with the levels in the

Netherlands and Finland.

Looking at trend developments, participation

rates have increased in all euro area countries

over the last decade, and to a greater extent than

in the preceding decade (for almost all countries

for which data are available for the 1980s 23). The

In the absence of more detailed data on immigration, it is not 21

possible to provide details on the explanation of this fi nding.

However, one might speculate that family reunifi cation and

integration issues and/or regulations governing access to the

labour market of non-EU workers may play a role.

A similar observation can be made for ethnic Germans 22

immigrating into Germany from former socialist countries. On

average, the unemployment rate of highly skilled ethnic German

immigrants (with the exception of engineers) is higher than that

of their medium-skilled counterparts. This can be explained,

at least partly, by a lack of applicability of existing skills. For

example, the human capital of lawyers depreciated immediately

and almost completely upon arrival.

No trend for the years 1984 to 1995 could be calculated for 23

Austria and Finland (where the EU-LFS started only in 1995) or

for Slovenia (fi rst EU-LFS collected in 1996).

Table 3 Overall participation rates in euro area countries

(employment rates in brackets)

Average annual change (percentage points) Level (%)Trend developments Recent developments

1984-1995 1) 1996-2007 2) 1996-2001 3) 2002-2007 4) 2007 5)

Belgium 0.2 (0.3) 0.4 (0.4) 0.3 (0.6) 0.5 (0.3) 66.7 (61.6)

Germany 0.3 (0.3) 0.4 (0.4) 0.1 (0.2) 0.7 (0.6) 75.6 (69.1)

Ireland 0.0 (0.1) 0.9 (1.2) 1.0 (1.8) 0.8 (0.6) 72.2 (68.9)

Greece 0.0 (0.0) 0.6 (0.6) 0.5 (0.3) 0.6 (0.8) 67.0 (61.5)

Spain 0.5 (0.3) 0.9 (1.6) 0.6 (1.8) 1.2 (1.4) 71.5 (65.8)

France -0.1 (-0.3) 0.2 (0.3) 0.2 (0.5) 0.2 (0.1) 69.6 (63.6)

Italy -0.1 (-0.3) 0.4 (0.7) 0.4 (0.6) 0.4 (0.7) 62.5 (58.9)

Luxembourg 0.0 (0.0) 0.4 (0.4) 0.6 (0.7) 0.2 (0.0) 65.6 (63.0)

Netherlands 0.9 (1.1) 0.8 (1.0) 1.1 (1.6) 0.5 (0.3) 78.5 (76.0)

Austria n.a. (n.a.) 0.2 (0.2) -0.1 (-0.1) 0.5 (0.4) 73.7 (70.3)

Portugal 0.0 (0.1) 0.5 (0.4) 0.7 (1.1) 0.3 (-0.2) 73.7 (67.6)

Slovenia n.a. (n.a.) 0.5 (0.6) 0.2 (0.4) 0.7 (0.8) 71.7 (68.3)

Finland n.a. (n.a.) 0.4 (1.0) 0.8 (1.6) 0.0 (0.4) 77.3 (71.3)

Euro area 0.2 (0.1) 0.5 (0.6) 0.4 (0.7) 0.6 (0.6) 70.8 (65.5)

Denmark 0.1 (0.3) 0.1 (0.3) -0.1 (0.3) 0.2 (0.2) 80.3 (77.3)

Sweden n.a. (n.a.) 0.2 (0.3) 0.1 (0.6) 0.3 (0.0) 79.9 (74.3)

United Kingdom 0.3 (0.4) 0.0 (0.2) 0.0 (0.5) 0.0 (0.0) 75.0 (71.1)

United States 0.3 (0.5) -0.1 (0.0) 0.0 (0.1) -0.3 (-0.2) 75.5 (72.0)

Sources: EU-LFS (spring data), ECB and NBB calculations.Note: 15 to 64 years old1) 1987-1995 for Spain and Portugal.2) 1997-2007 for Slovenia and 1996-2006 for the United States.3) 1997-2001 for Slovenia.4) 2002-2006 for the United States.5) 2006 for the United States.

21ECB


June 2008

3 MAIN TRENDS IN

LABOUR SUPPLYNetherlands outperformed all other euro area

countries, moving from the lowest participation

rate in 1983 to the highest rate in 2007. During

the past decade, the increase in participation

was strongest in Spain and Ireland. In other

countries, like Luxembourg, Italy and Belgium,

which also had low participation levels in 1996,

the advance was more limited.

However, over the last decade, the apparent

acceleration of participation rates for the

euro area as a whole is not widespread across

individual countries. In six countries (Belgium,

Germany, Greece, Spain, Austria and Slovenia)

the increase in participation was clearly higher

in 2002-07 than between 1996 and 2001.

However, participation growth rates more or less

stabilised in France and Italy and decelerated in

the remaining fi ve countries.

Although remaining lower than for males,

female participation is especially high in Finland

and the Netherlands, but low in Italy, Greece

and Luxembourg (see Annex 3, Table 28).

Low-skilled persons participate substantially

more often in Portugal, and only to a relatively

limited degree in Belgium and Italy. Together

with Spain, Portugal also registers the highest

participation rates of non-EU-citizens. The latter

group participates only to a rather limited degree

in Belgium and the Netherlands.

3.2.3 COMPOSITIONAL EFFECTS IN PARTICIPATION

Although over time, there has been a broad-

based increase in participation across the various

sub-groups considered, the observed evolution

of the euro area aggregate participation rate

can in part be explained by the effects of

changes in the composition of the working

age population.24 Ceteris paribus, this factor

appears to explain almost half of the observed

increase in participation over the last decade 25

(see Table 2), mainly as a result of the gradual

rise in average education level and the shift in

the age structure of the population (the share

of young people in the working age population

decreased in favour of the 25-54 year olds). But

during the most recent fi ve-year period, when

the participation rate increased on average by

0.6 percentage point per year, the impact of

changes in composition fell to one-fourth. Thus

the larger contribution (0.4 percentage point)

to the increase in labour market participation

over the last fi ve years is attributable to a real

underlying increase in participation behaviour.

For individual countries, over the last decade

the contribution of the population structure

effect to participation 26 was especially important

in Italy and Greece. For these countries, the

population effect explained more than three

quarters of the observed participation rate

increase. In Belgium and Ireland it accounted

for about half of the increase, and in Austria,

Finland and France its impact was limited to

one-fi fth or less. In these latter three countries,

the positive impact of the rise in education level

was partly offset by a negative impact from the

ageing of the population towards the 55-64 age

group. During the 2002-07 period, the latter

phenomenon became more widespread,

implying that the population effect was in

general more limited across euro area countries.

In addition to infl uencing the development of

the overall participation rates, differences in the

population structure also have an impact on the

observed participation rate level across euro

area countries.27 As can be seen in Chart 2 for

the year 2007, the impact of the population

As participation structurally depends on factors like gender, age 24

and education level, a shift in the relative shares of these groups

in the population affects the development of the observed overall

participation rate.

For this aim, detailed EU-LFS data on the participation rates 25

of 18 subgroups of the population of working age – subdivided

according to gender (men, women), age (15-24, 25-54, 55-64)

and education level (low, medium, high) – were weighted by

using the population composition of 1995. As (the evolution

of) participation also differs substantially according to

nationality, ideally this factor should also be taken into account.

Unfortunately, for the 1990s, EU-LFS data for several euro area

countries do not provide this subdivision. This makes such a

decomposition exercise unreliable.

Due to a lack of data, no reliable results are available for 26

Germany, Spain, Luxembourg, the Netherlands and Slovenia, or

for Denmark, Sweden and the United Kingdom.

Indeed, for two countries differing only in population structure, 27

the country with, for instance, proportionally more national

highly skilled men aged 25 to 54 will have a higher observed

overall participation rate than the country where the population

proportionally consists of more young or more older, low-skilled

non-EU women.

22ECB


June 2008

structure is particularly important for some

countries.28 The effect is most pronounced for

Slovenia, where the observed participation rate

is close to the euro area average, but, having a

population with almost no foreigners and

relatively few low-skilled persons, its rate would

be clearly below this average if it were to have

the same population composition as the euro

area as a whole. Also in countries such as

Luxembourg, the Netherlands and Finland, the

positive effect of population composition is

important, mainly due to a relatively high-

skilled population.29 Conversely, the

participation rates in Italy and Portugal are

considerably higher if one corrects for the

population structure effect, as, according to the

EU-LFS, the population of both countries is on

average less skilled than in the euro area as a

whole. Excluding this population structure

effect, the range of participation rates among

euro area countries is reduced, but still remains

large. The Netherlands still has the highest

participation rate among euro area countries,

while the lowest adjusted rates are found in Italy

and Luxembourg.

3.2.4 COHORT EFFECTS IN PARTICIPATION 30

This section considers to what extent trend

developments in the euro area participation

rate can also be attributed to so-called cohort

effects. These effects are due to differences in

labour market participation across birth cohorts

emerging as a result of different individual

participation choices made early in life

(e.g. regarding fertility, maternity leave and/

or education) and persisting throughout the

life-cycle. Beyond potential crowding-out

effects stemming from the size of the cohort

entering the labour market and differences

in human capital investment over time, these

cohort effects are likely to refl ect evolving

preferences, social norms and/or institutions.31

Chart 3 shows participation rates by age group

of different cohorts in the euro area population.

A cohort refers to persons in the same age

group, thus persons born within a particular

Due to detailed data availability concerning nationality in the 28

EU-LFS for 2007, it was possible to do this decomposition exercise

using 54 population groups, obtained using the combination

of the following factors: gender (men, women), age (15-24,

25-54, 55-64), education level (low, medium, high) and nationality

(nationals, other EU-citizens, non-EU-citizens). To exclude the

population structure effect, the country specifi c participation rates

of these groups were weighted by using the population structure

of the euro area as a whole. For Spain, Ireland, Denmark, Sweden

and the United Kingdom, the results suffer from reliability

problems, as too large a proportion of the population was not

subdivided according to the four factors used.

For Luxembourg, this exercise is less conclusive, as currently 29

38% of persons working in the country are commuters. They

reside in one of the neighbouring countries, and are therefore

part of the population of those countries. As the EU-LFS data

for Luxembourg consider only the Luxembourg population, they

do not refl ect the situation of the whole labour force.

Prepared by A. Balleer and J. Turunen.30

So-called cohort effects generally encompass any factor 31

associated with a particular birth year, e.g. general economic

conditions or crowding-out effects. Empirical evidence suggests

that the size of the cohort entering the labour market greatly

affects participation. There are many explanations for this,

e.g. depressed earnings (see Welch (1979), Berger (1985) and

Korenman and Neumark (1997)), also formulated as the “relative

income” hypothesis, i.e. large cohorts experience lower incomes

relative to their expectations (Easterlin 1980). Finally, large

events like WWII may result in cohort effects as Acemoglu et

al. (2002) fi nd evidence that, owing to men going to war, more

women worked and also stayed in the labour market in the US.

Chart 2 Overall participation rate in 2007: correction for the population structure effect 1)

(differences, in percentage points, with respect to the euro area average)

-10

-8

-6

-4

-2

0

2

4

6

8

10

-10

-8

-6

-4

-2

0

2

4

6

8

10

observed

adjusted 2)

1 Belgium

2 Germany

3 Ireland

4 Greece

5 Spain

6 France

7 Italy

8 Luxembourg

9 Netherlands

10 Austria

11 Portugal

12 Slovenia

13 Finland

14 Denmark

15 Sweden

16 United Kingdom

17 United States

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

Sources: EU-LFS (spring data) and NBB calculations.Note: 15 to 64 year olds.1) For Spain, Ireland, Denmark, Sweden and the UK the adjusted participation rate is not displayed, as the mathematical impact originating from the fact that the population used excludes the EU-LFS respondents for which not all characteristics concerning gender, age, education level and nationality are available, was too large. For the United States, no detailed data available to calculate an adjusted participation rate.2) Remaining difference, after correction for data non-availability and population structure (composition of the population by gender, age, education level and nationality).

23ECB


June 2008

3 MAIN TRENDS IN

LABOUR SUPPLY

time period, and is represented by a separate

line in the graph.32 The cohort effect is

measured as the vertical distance in

participation rates between the different cohorts

for a given age. The Charts suggest a substantial

cohort effect for females, but no visible cohort

effects for males. Female participation has

risen by close to 20 percentage points when

comparing the participation rate of the

1943-1947-born cohort with the 1963-1967-

born cohort at the age of 40-44 years old.33 In

addition to the substantial level effect, the

shape of the profi le between ages 20 and 35

changes for the younger cohorts as the kink

that is visible in the profi le of those born

1963-1967 at the ages 30-34 disappears. The

timing of this effect suggests that the

differences across cohorts may refl ect a number

of factors relating to, e.g., improved

possibilities for reconciling family and

employment, and shifts towards postponed

motherhood or longer education. Potentially

due to the latter effect, participation of persons

aged 15-24 has declined between the cohort

born in 1968-1972 and the two youngest

cohorts for both males and females.

Table 4 presents the vertical distance

(i.e. the difference in participation levels)

between the youngest and the oldest cohort

for females between the ages of 25 and 44 in

individual countries over the recent decade.

The increases in levels of female participation

between the youngest and older cohort are

particularly large in Ireland, Greece, Spain,

Luxembourg and the Netherlands, but relatively

small in Austria and France. (For comparison,

the changes are even smaller in the UK and

It is generated by starting in the year 2005 for a particular 32

age group, and tracking the participation rate of each cohort

backwards in time over fi ve-year intervals. It has to be noted

that this procedure does not identify a pure cohort effect, but

rather its interaction with age (a so-called age-cohort). Owing to

the relatively short time series dimension, cohort profi les only

partially overlap.

This is the vertical distance between the participation rate of the 33

group born between 1943 and 1947 and the group born between

1963 and 1967 in Chart 3.

Chart 3 Age-cohort profiles in the euro area

(1983-2007)

born 1943-1947

born 1948-1952

born 1953-1957

born 1958-1962

born 1963-1967

born 1968-1972

born 1973-1977

born 1978-1982

y-axis: participation rate

x-axis: age groups

Females Males

10

20

30

40

50

60

70

80

90

10

20

30

40

50

60

70

80

90

1

15-19

20-24

25-29

1

2

3

30-34

35-39

40-44

4

5

6 55-59

45-49

50-54

7

8

9

2 3 4 5 6 7 8 910

20

30

40

50

60

70

80

90

100

110

10

20

30

40

50

60

70

80

90

100

110

15-19

20-24

25-29

1

2

3

30-34

35-39

40-44

4

5

6 55-59

45-49

50-54

7

8

9

1 2 3 4 5 6 7 8 9

Sources: LFS and ECB calculations.

24ECB


June 2008

Sweden and even negative in Denmark.) It has

to be noted that for the ages discussed above,

but also for the very young workers, the shape

of the participation profi les, as well as the levels

reached in 2007, are very heterogeneous across

countries. This may be linked, among many

other reasons, to differences in investment

in education or educational systems across

countries.

The cohort profi les shown in Chart 3 are

potentially infl uenced by both business cycle

factors and evolving institutions (which may

also help to explain the country differences).34 35

Overall, a continued increase in the proportion

of women in the labour force and demographic

changes which shift the relative share of the

labour force between birth cohorts imply that

cohort effects will continue to affect euro area

labour supply in the future. Box 1 considers

future developments in labour supply further.

See e.g. Genre et al. (2007) for a study on the role of institutions 34

for participation rates by age and gender. In this study, lagged

participation tries to capture cohort effects for older women

(those between 55-64). This term is statistically signifi cant.

Estimating a cohort-based model of participation for the euro 35

area and most euro area countries using the model presented in

Aaronson et al. (2006) and Fallick and Pingle (2007) which controls

for business cycle and age effects, confi rms that cohort effects are

statistically signifi cant for females and are robust to period effects as

measured by an indicator of the business cycle, while cohort effects

for males are not statistically signifi cant for the euro area. The

estimation suggests a negative effect on participation for the female

cohorts born between 1922 and 1940 (entering the labour market

around the early 1940s and early 1960s) and a positive effect for

those born between 1966 and 1973 (entering in the late 1980s and

early 1990s). The higher participation-propensity of females born

in the late 1960s and early 1970s has therefore contributed to the

increase in female participation in the euro area. The overall pattern

of increasing cohort effects for women appears similar to that

observed in the United States, see Figure 8 in Fallick and Pingle

(2007). In addition, the results suggest that the cycle has strongly

infl uenced the participation rates of the youngest cohorts.

Table 4 Participation differences for females between cohorts in single countries. Change in participation from 1997 to 2007

(percentage points and levels in 2007)

Country Age group 25-29 30-34 35-39 40-44

1997-2007 2007 1997-2007 2007 1997-2007 2007 1997-2007 2007

Belgium 7.4 84.7 8.2 80.7 17.7 81.1 24.5 80.2

Germany 7.2 78.2 13.0 80.8 14.4 81.5 16.7 83.8

Ireland 19.6 82.4 29.2 75.1 34.7 70.3 34.6 67.9

Greece 24.4 77.8 19.2 73.4 21.7 74.4 23.4 71.6

Spain 21.7 80.6 29.9 78.1 36.3 73.8 26.3 72.8

France 5.6 81.2 8.1 80.2 10.8 82.8 12.2 84.2

Italy 2.8 62.7 11.4 68.9 11.7 67.1 16.9 64.7

Luxembourg 9.9 75.1 25.6 80.4 23.1 72.3 25.6 68.0

Netherlands 20.5 86.1 30.5 84.7 25.1 81.6 29.6 81.9

Austria -3.3 76.7 1.8 78.0 6.9 83.4 7.4 85.9

Portugal 12.5 84.9 13.9 87.9 19.8 87.6 14.5 84.7

Slovenia 3.3 79.0 -1.2 81.1 -1.8 85.2 1.1 89.7

Finland 9.6 77.4 14.2 77.9 16.6 77.7 20.8 78.1

Denmark -5.1 82.4 -5.2 84.3 -1.4 86.8 -3.5 85.4

Sweden 2.6 83.5 4.9 87.8 1.5 88.4 -0.1 88.8

United Kingdom 10.8 76.5 10.1 74.9 5.8 76.5 4.1 79.3

United States 1) 0.2 74.9 -0.7 74.1 -2.2 74.0 -1.0 77.0

Sources: LFS and ECB calculations. 1) Developments for the United States are 1995-2007, information is taken from Aaronson et al. (2006) for 1995 and from the Bureau of Labour Statistics for 2007.

25ECB


June 2008

3 MAIN TRENDS IN

LABOUR SUPPLYBox 1

PROJECTIONS OF LABOUR SUPPLY IN THE EURO AREA AND EURO AREA COUNTRIES 1

Demographic developments

A key determinant of labour force developments

is the underlying demographic trend. In

particular, the evolution of birth rates gives an

early indication of future population structure.

Chart A shows that birth rates in the euro area

tended to follow a downward trend from the

mid 1960s until the mid 1990s, remaining

broadly stable thereafter. This general trend

holds for most euro area countries, with

considerable heterogeneity in fertility rate

levels across countries (as indicated by the

distance between the maximum and minimum

birth rates). Ireland and France emerge as the

countries with the highest birth rates in recent

years, and Germany, Italy and the Netherlands

are among those with the lowest. The observed

thirty-year decline in birth rates should lead to

a prolonged decline in the share of prime-aged workers in the total population in the near future.

Since this group exhibits the highest participation rate in the labour market (see Section 3.2.1),

this should also translate into a slowdown in the growth of the labour force.

Future participation rates and labour force developments

On the basis of population projections and assumptions about group-specifi c participation rates,

it is possible to make projections of labour force developments in the euro area.2

According to the latest Eurostat population projections, the working age population is expected

to grow slightly up to 2011 and then decline (although at varying intensity) across the rest of

the forecast horizon (see Table). Under the assumption that participation rates by gender and

age groups remain constant at the 2007 level, the euro area labour force would shrink by an

average of 0.5% per year over the period 2007-2050. This refl ects the continued ageing of

the working age population and, therefore, the previously mentioned decline in the share of

prime-aged workers. Indeed, the overall participation rate would fall by 1.5 percentage points

to 69.4% in 2050 (see Table). As a result of population ageing, the old-age dependency ratio

1 Prepared by R. Gomez-Salvador and A. Novo.

2 Population projections are based on Eurostat’s “baseline variant” for the working age population on January 1st of each year, based on

assumptions on fertility, life expectancy and net migration. In particular, the projections are based on the assumption that the share of

the net migration over the total population remains stable at 0.4% over the forecast horizon. The labour force developments are then

obtained following the methodology described in Shimer (1998).

Chart A Crude birth rates in the euro area 1), 2)

5

10

15

20

25

5

10

15

20

25

2005

average

max

min

1960 1965 1970 1975 1980 1985 1990 1995 2000

Sources: Eurostat and ECB calculations. 1) Ratio of the number of births during the year, per 1,000 inhabitants. 2) Average, maximum and minimum birth rates across euro area countries.

26ECB


June 2008

(i.e. the elderly population divided by the working age population) is expected to increase from

around 25% in 2007 to around 55% in 2050.3, 4

For euro area countries, the impact of population structure on the future overall participation rate

is negative in almost all cases, but to varying degrees. Together with the general decline in the

working age population, this will lead to a decline in the future labour force in most countries.5

An alterative scenario

Based on a European Commission ageing report,6 a second scenario for the euro area can be

considered. Relative to the baseline scenario, which assumes that group-specifi c participation

rates remain stable at their 2007 level, the alternative scenario projects higher participation

rates after 2007, particularly among women, whose participation rates have been increasing

over recent decades, and older workers, due to recently enacted public pension system reforms

(see also Section 4.1.3 and Box 6). Under this scenario, the euro area labour force would contract

by 0.3% per year on average over the period 2007-2050, that is, by less than in the previous

scenario. Indeed, the overall participation rate in this alternative scenario would increase by

4.2 percentage points to 75.1% in 2050, to around 5 percentage points higher than in the scenario

presented above. However, the positive contribution from increased labour market participation

would only result in positive developments in the labour force until 2015 (the labour force

growing by 0.4% per year on average). From that year onwards, the negative contribution coming

3 It is worth mentioning that two periods can be distinguished with regard to the impact of ageing on the participation rate. From 2007 to

2030 the participation rate is projected to fall by around 2.5 p.p to 68.5%, and then, as the share of prime-aged workers starts to increase

again, to recover by around 1 p.p in 2050 (see Table). However, the positive contribution coming from this increased participation in

the last part of the projection horizon is not expected to have any signifi cant impact on labour force developments, being outweighed by

the continued decline in the working age population (see Chart B).

4 For detailed country-specifi c and euro area charts on the projected developments in dependency ratios, see Maddaloni, et al (2006).

5 The only two exceptions are Ireland and Luxembourg, where positive developments in the working age population may more than

offset the decline in participation, and therefore translate into an increase in the labour force (see Table).

6 European Commission (2006, 2007). This second scenario does not assume a general rise in education levels, but analyses the effects

of expected demographic and labour market developments given the present enrolment and cost situation.

Working age population, participation rate and labour force developments 1)

Working age population 2) Participation rate 3) Labour force 2)

2007-10 2011-30 2031-50 2007-50 2007 2010 2030 2050 2007-10 2011-30 2031-50 2007-50

Belgium 0.3 -0.3 -0.2 -0.2 66.7 66.1 65.2 65.3 0.0 -0.4 -0.2 -0.3

Germany -0.1 -0.6 -0.7 -0.6 75.6 75.9 74.4 74.5 -0.1 -0.7 -0.7 -0.6

Ireland 0.8 0.6 -0.2 0.3 72.2 72.0 70.3 70.8 0.9 0.5 -0.2 0.2

Greece 0.2 -0.3 -1.0 -0.6 67.0 67.0 63.6 64.8 0.2 -0.6 -0.9 -0.6

Spain 0.2 -0.2 -1.2 -0.6 71.5 71.3 67.0 68.7 0.2 -0.6 -1.1 -0.7

France 0.3 -0.2 -0.2 -0.1 69.8 69.2 68.3 68.7 0.1 -0.2 -0.1 -0.2

Italy -0.1 -0.5 -1.0 -0.7 62.5 62.3 58.6 60.3 -0.2 -0.8 -0.9 -0.8

Luxembourg 1.0 0.5 0.5 0.5 65.6 64.6 63.2 63.2 0.6 0.4 0.5 0.4

Netherlands 0.3 -0.2 -0.1 -0.1 78.5 78.2 77.4 78.0 0.1 -0.2 -0.1 -0.1

Austria 0.3 -0.3 -0.5 -0.4 74.9 74.6 72.0 72.2 0.2 -0.5 -0.5 -0.4

Portugal 0.0 -0.4 -0.9 -0.6 73.7 73.8 71.1 72.0 0.0 -0.5 -0.9 -0.6

Slovenia 0.0 -0.6 -0.8 -0.6 71.7 71.8 69.0 69.4 -0.1 -0.8 -0.7 -0.7

Finland 0.2 -0.5 -0.2 -0.3 77.3 76.8 77.5 77.2 0.0 -0.5 -0.3 -0.3

Euro area 0.1 -0.4 -0.7 -0.5 70.8 70.8 68.5 69.4 0.0 -0.5 -0.6 -0.5

Sources: Eurostat and ECB calculations. 1) Working age population derived from Eurostat projections (baseline scenario). Overall participation rate derived by keeping participation rates by gender and age group constant at the 2007 level. 2) Annual growth rates. 3) Levels expressed as percentages.

27ECB


June 2008

3 MAIN TRENDS IN

LABOUR SUPPLY

3.3 HOURS WORKED 36

To evaluate labour utilisation, it is important to

look at not only the (relative) number of people

participating in the labour market (the extensive

margin), but also the number of hours worked per

employed person (intensive margin). Indeed, in

the context of population ageing and shrinkage

of the total labour force, both channels can be

used to increase the effective labour supply.

To obtain a fuller picture of labour supply, this

section considers average yearly hours of work

at an aggregate level from 1991 onwards using

data from the OECD.37 Unfortunately these data

are not available for different sub-groups of the

labour force, so this analysis uses EU-LFS data

on usually-worked weekly hours.38 The data on

both concepts are comparable, since the OECD

uses the EU-LFS data as input for calculating

annual hours.39 In addition, data on annual

and weekly working hours are closely linked

(see Chart 12 in Annex 3).

3.3.1 HOURS WORKED IN THE EURO AREA AND IN

THE UNITED STATES

The main difference between hours of work in

the United States and Europe is the number of

hours worked in annual terms. According to the

most recent OECD data, an average euro area

worker worked 1,672 hours in 2005,40

Prepared by J. De Mulder and R. Gomez-Salvador.36

See Annex 2 for a short description of this dataset.37

The EU-LFS also provides information on actual weekly 38

working hours, but this concept is – more than the usual working

hours – infl uenced by one-off factors (for instance exceptional

absences or overtime work) during the reference week of the

survey.

The OECD fi gures on hours worked per year are obtained by 39

combining EU-LFS information on weekly hours (usual hours,

overtime work and hours on additional jobs) with the number of

weeks worked per year.

As Slovenia is not an OECD member, there are no OECD data 40

available for the euro area as a whole. Therefore a euro area

average of the other twelve countries was calculated, weighted

by employment (number of persons at work). Before 1995,

no data are available for Austria and Finland, so the euro area

fi gures for the period 1991-1994 correspond to the weighted

average of ten countries.

from demographic developments would dominate, and labour force growth would start declining

by 0.5% per year on average until 2050 (see Chart C).

In sum, even if the most dominant recent trends in labour market participation, such as the

increase in female participation continue, the labour force can be expected to decline in the near

future, due to the projected fall in the working age population.

Chart B Euro area labour force projections – constant participation rates 1)

Chart C Euro area labour force projections – varying participation rates 2)

labour force (left-hand scale)participation rate (right-hand scale)

100,000

110,000

120,000

130,000

140,000

150,000

160,000

170,000 76

62

64

66

68

70

72

74

1983 1995 2007 2016 2025 2037100,000

110,000

120,000

130,000

140,000

150,000

160,000

170,000

1983 1995 2007 2019 2031 204362

64

66

68

70

72

74

76

labour force (left-hand scale)

participation rate (right-hand scale)

Sources: Eurostat and ECB calculations. 1) Constant participation rates across gender and age groups (15-24, 25-29, 30-34, 35-39, 40-44, 45-49, 50-54, 55-59 and 60-64) at the 2007 level. 2) Varying overall participation rates at the country level.

28ECB


June 2008

considerably less than the 1,922 hours worked in

the United States. This suggests that regulations

on hours of work and annual holidays may play

an important role in explaining the differences

between annual and weekly hours in the United

States and euro area. Over the past decade,

average annual hours of work also fell faster in

the euro area than in the United States (by

respectively 0.3% and 0.1% per worker per year).

However, since 2003, a minor increase in annual

hours has been observed in the euro area (see

Chart 4).41 With regard to weekly hours, EU-LFS

data show that the decline in average working

time per worker had already started in the early

1980s, and the reduction in hours of work in

Europe in recent years has come mainly from a

relative decline in hours per week. The average

working week was 40.1 hours in the euro area in

1983, decreasing to 37.4 hours per week in 2007.

Comparable data for the United States are only

available for the period 1996-2005. The data

show that the average working week in the United

States was around 39 hours in 1996, comparable

to hours worked in the euro area. However, since

then the length of the average working week has

roughly stabilised in the United States.42 In 2005,

the latest available year, it was 38.5 hours.

The information on annual hours worked per

worker can be combined and compared with

data on the number of people employed to allow

a qualifi cation of the developments in total hours

worked in the euro area (see Chart 5). This shows

that following a slight decline, both the number

of people employed and total hours worked show

a similar upward trend since the mid 1990s,

although employment increased faster than

total hours worked (while total hours appears to

have been more pro-cyclical).43 Moreover, the

picture of employment developments in the last

decade changes somewhat when labour input is

measured in hours (rather than in the number of

people). It appears that instead of the decline

in average growth rates between 1996-2001

and 2002-07 recorded by the employment rate

(as shown in Table 1), the growth in total hours

has remained broadly stable (at 1.1%), similar to

the growth rate of total hours in the Unites States.

Finally, it is worth mentioning that this different

pattern has an impact on measured productivity

growth in the euro area, which shows a less

marked slowdown in the second half of the

This is driven by the smaller number of annual weeks worked 41

in Europe, fewer overtime hours worked on the main job in

Europe, and fewer hours on additional jobs in Europe.

According to the Groningen Growth and Development Centre 42

(GGDC), hours worked were also stable in the US over the

whole of the 1980s and 1990s.

Indeed, looking at the period 2001-2004, it appears that hours 43

worked per worker have acted more as a buffer to economic

conditions, given that total hours declined slightly and

then recovered, in line with the economic slowdown, while

employment growth remained slightly positive.

Chart 4 Hours worked per worker in the euro area and the United States

(euro area (EA) vs. United States (US))

2,000

1,900

1,800

1,700

1,600

1,5001983 1987 1991 1995 1999 2003 2007

46

44

42

40

38

36

annual hours EA (left-hand scale)

annual hours US (left-hand scale)

weekly hours EA (right-hand scale)

weekly hours US (right-hand scale)

Sources: Eurostat, OECD and ECB calculations.

Chart 5 Developments in employment, total hours worked, and productivity per person and per hour in the euro area

88

92

96

100

104

108

112

1991 1993 1995 1997 1999 2001 2003 2005-2.0

-1.0

0.0

1.0

2.0

3.0

4.0

total hours - 1991=100 (left-hand scale)

employment - 1991=100 (left-hand scale)

productivity per hour (right-hand scale)

productivity per person (right-hand scale)

Sources: OECD, EU-LFS and own calculations. Employment in thousands and Total hours in millions.

29ECB


June 2008

3 MAIN TRENDS IN

LABOUR SUPPLY1990s when measured in hours worked instead

of number of employed persons.

3.3.2 HOURS WORKED IN THE EURO AREA BY

DIFFERENT WORKER GROUPS 44

This section examines whether the number

of hours worked per worker in the euro area

varies by worker group. In addition to the

characteristics of gender, age, education level

and nationality, professional status (employee

or self-employed) and work regime (part-time

or full-time work) may also affect average hours

across countries.45

An important explanation for the downward

trend in euro area hours worked per worker per

week is the substantial increase in part-time

work.46 As can be seen in panel B of Table 5,

the average part-time worker worked 20 hours

per week in 2007, compared with 41.5 hours

for a full-time worker. In 1983, only 9% of

the working age population was working part-

time; however, by 2007, this fi gure had risen to

19%. This development is linked to the rise in

female participation described in the previous

section (see also Section 4.2). In 2007, 35%

of all working women had a part-time job,

compared with only 7% of men; thus, the greater

importance of part-time work for women also

explains their lower average number of hours

worked.

At an average of 34 hours per week in 2007,

young persons worked the fewest hours. This

fi gure mainly refl ects the fact that many young

people work part-time as they combine their

studies with a job.47 There is little difference

Prepared by J. De Mulder. 44

The latter two factors were not treated in the section on 45

participation rates, as they only concern the working population,

whereas the labour force is composed of both the working and

the unemployed.

In the EU-LFS, the full-time versus part-time distinction is based 46

on the declaration by the respondent, except in the Netherlands,

where this subdivision is made according to whether the

respondent usually works at least 35 hours (full-time) or less

(part-time).

In 2006 (last year for which this information is available), an 47

average working student aged 15 to 24 worked for roughly

13 hours, which has a considerable downward infl uence on the

average number of hours worked by young persons.

Table 5 Hours worked per worker in the euro area

Panel A: Total

Average annual change (%) Level (hours) Trend developments Recent developments

1984-1995 1996-2005 1996-2001 2002-2005 2005

Annual hours n.a. -0.3 -0.3 -0.1 1,672

1984-1995 1996-2007 1996-2001 2002-2007 2007

Weekly hours -0.3 -0.3 -0.3 -0.2 37.4

Excluding the effect of changes in the population composition 1) n.a. -0.2 -0.3 -0.2Panel B: Weekly hours in 2007 according to different subdivisions (hours)According to work regime According to gender

Full-time work 41.5 Males 40.9

Part-time work 20.0 Females 32.9

According to age According to education level15-24 years old 33.7 Low 38.2

25-54 years old 37.8 Medium 37.1

55-64 years old 37.3 High 38.4

According to nationality According to professional statusNationals 37.4 Employees 35.9

Other EU15-citizens 36.9 Other workers 45.5

Non EU15-citizens 36.1

of 12 new EU member states 37.3

of non EU27-countries 35.3

Sources: EU-LFS (spring data), ECB and NBB calculations. Note: 15 to 64 years old, except for the subdivision according to education level (25-64 years old). 1) Calculated by weighting the hours worked by 36 subgroups of the population of working age (subdivided according to gender, age, education level and professional status) with the structure of the population of working age in 1995.

30ECB


June 2008

in hours worked per week between prime-aged

and older workers, according to education level

and nationality. In contrast, there is signifi cant

variation in the average hours worked by

professional status. In 2007, employees worked

on average 36 hours per week, while the average

week was 45.5 hours long for other workers

(mainly self-employed).

The decline of weekly working time since 1983

was broadly based and comparable for most

worker groups. Nevertheless, the decrease was

somewhat stronger for females, due to the faster

up-take of part-time work, and for the younger

and older generations. The latter developments

most likely refl ect the larger proportion of

young people pursuing tertiary education and

the higher participation rate of older workers,

who remain at work longer in life but reduce

their average weekly working time.

3.3.3 HOURS WORKED PER WEEK IN THE EURO

AREA COUNTRIES

As with labour market participation rates,

signifi cant variation in the average yearly hours

of work is apparent across euro area countries

(see Table 6).48 In 2005, average working time

ranged from almost 2,000 hours per worker per

year in Greece, to some 1,400 hours in the

Netherlands.49 In most other countries, the

average person worked between 1,600 and

1,800 hours (compared with 1,600 hours in

Denmark, Sweden and the United Kingdom).

During the past decade (until 2005), developments

As mentioned before, no data for yearly hours worked are 48

available for Slovenia.

These observed differences are largely attributable to differences in 49

the composition of employment, mainly as regards the occurrence

of part-time work and self-employment. Indeed, if only full-time

employees are considered, average working hours in Greece only

exceed those in the Netherlands by some 100 hours.

Table 6 Hours worked in euro area countries

Annual hours 1) Weekly hours Average annual change (%) Level

(hours) Average annual change (%) Level

(hours) Trend Recent developments Trend developments Recent developments

1996-2005 1996-2001 2002-2005 2005 1984-1995 1996-2007 1996-2001 2002-2007 2007

Belgium 0.0 0.0 0.1 1,681 -0.5 -0.1 -0.1 -0.2 37.1

Germany -0.4 -0.4 -0.6 1,622 -0.5 -0.5 -0.4 -0.6 35.6

Ireland -0.9 -1.3 -0.2 1,729 -0.5 -0.9 -1.3 -0.6 36.3

Greece -0.1 -0.1 0.0 1,995 -0.4 -0.2 -0.1 -0.3 42.5

Spain 0.0 -0.1 0.0 1,791 -0.3 -0.2 -0.2 -0.3 39.3

France -0.6 -0.6 -0.6 1,592 -0.4 -0.1 -0.7 0.5 38.0

Italy 0.1 -0.2 0.5 1,730 -0.1 -0.2 -0.2 -0.2 38.5

Luxembourg -0.7 -0.6 -0.9 1,637 -0.2 -0.6 -0.5 -0.7 36.7

Netherlands -0.2 -0.3 0.0 1,417 -1.1 -0.5 -0.7 -0.4 30.9

Austria 0.2 0.1 0.3 1,729 n.a. 0.0 -0.2 0.3 39.0

Portugal -0.8 -1.3 -0.1 1,775 -0.5 -0.6 -0.9 -0.2 39.6

Slovenia n.a. n.a. n.a. n.a. n.a. -0.4 -0.1 -0.6 40.3

Finland -0.1 0.0 -0.4 1,682 n.a. -0.1 0.1 -0.2 37.9

Euro area -0.3 -0.3 -0.1 1,672 -0.3 -0.3 -0.3 -0.2 37.4

Denmark 0.3 0.8 -0.3 1,586 -0.3 0.0 0.3 -0.3 35.7

Sweden 0.5 0.9 -0.2 1,558 n.a. 0.1 0.2 -0.1 36.7

United Kingdom -0.4 -0.2 -0.7 1,662 0.0 -0.3 -0.3 -0.3 37.2

United States 2) -0.1 -0.2 0.0 1,922 n.a. -0.1 -0.1 -0.1 38.5

Sources: EU-LFS (spring data) and OECD calculations. Note: 15 to 64 years old. 1) OECD annual hours data not available for the 1980s, so the trend development for 1984-1995 could not be calculated. 2) For annual and weekly hours in the United States, the considered trend period and the fi rst recent development period start in 1997 instead of 1996. For weekly hours, the second recent development period concerns 2002-05, and the level in the last column concerns 2005.

31ECB


June 2008

3 MAIN TRENDS IN

LABOUR SUPPLYhave diverged substantially across countries.

While average working hours rose in Austria,

considerable decreases were recorded in Ireland,

Portugal, Luxembourg and France recorded

considerable decreases. Since 2002, the

downward trend has continued in the latter two

countries. In other countries, especially Italy,

average annual working hours actually increased.

The order of the euro area countries according to

the average number of weekly hours worked is

roughly comparable to the one based on average

annual hours. In 2007 average weekly working

time was highest in Greece (42.5 hours), and

by far lowest in the Netherlands (30.9 hours).

In Denmark, Sweden and the United Kingdom,

the situation was comparable to the euro area

average, with 36 to 37 hours.

Average weekly hours worked fell in almost all

European countries over all considered periods;

only in Austria and France was an increase

observed during the last fi ve-year period. The

decrease was strongest in the Netherlands

(especially in the 1984-1995 period) and in

Ireland (during the last decade). An important

factor in both the level and downward trend in

average weekly hours worked is the rate of

part-time work, which may be linked to both

institutional features and individual preferences.

For example, Greece, which has the highest

average working week, also experiences one of

the lowest rates of part-time work among the

euro area countries. Similarly, the Netherlands,

with the shortest working week, has the highest

rate of part-time work. Furthermore, the

decrease in the average hours worked per week

in the Netherlands coincided with a strong

increase in part-time work following changes to

part-time work legislation in 1982.50

3.3.4 COMPOSITIONAL EFFECTS IN HOURS

WORKED

The overall impact of the evolution of the

composition of employment on observed

working hours 51 is rather limited for the euro

area as a whole (see panel A of Table 5). Overall,

the downward infl uence of the higher proportion

of females and employees in employment was

partly compensated by a decreasing share of

young workers.

Contrary to the euro area average, changes in

the composition of the employed population

seem to have had a considerable impact in some

countries.52 In Belgium, Spain, Finland, Greece

and Italy, they accounted for almost the entire

development of average working hours over

the last decade, implying that the underlying

working hours of the various subgroups

have hardly changed. In France, a downward

development took place, which was partly

counteracted by changes in the population

structure. In most other countries, the effects of

changes in the population structure have been

limited over the last fi ve years.

Part of the observed differences in average hours

worked between countries in 2007 can also be

attributed to the different country-composition

of employment.53 For example, the short

working week in the Netherlands can partly be

explained by a relatively larger share of young

workers in employment than on average in the

euro area, while part of the longer working week

in Greece can be attributed to a relatively high

proportion of male workers and self-employed.

For a discussion of the explanations for the differences in hours 50

of work across countries, see also Leiner-Killinger, Madaschi

and Ward-Warmedinger (2005).

For this ceteris paribus exercise, detailed EU-LFS data 51

concerning the hours worked of 36 subgroups of the working

age population – subdivided according to gender (men, women),

age (15-24, 25-54, 55-64), education level (low, medium, high)

and professional status (employee or self-employed) – were

weighted using the employment composition of 1995. In the

previous section, these four factors were identifi ed as the most

important distinguishing characteristics concerning hours

worked in the euro area. The work regime was not taken into

account, since defi nitions of part-time and full-time work depend

on the national legal system (for instance, a person working

38 hours per week is working (more than) full-time in France

and Belgium, but is working part-time in other countries). In

addition, this variable is closely linked to gender.

The size of the population effect depends on the relative size 52

of the different evolutions inside the employed population. In

most countries, the share of females and employees is growing,

having a downward effect on average hours worked. On the

other hand, the proportion of young persons in employment is

decreasing, pushing upwards the average working time.

Again, detailed EU-LFS data, subdivided according to gender, 53

age, education level and professional status, were used. To exclude

the population structure effect, the country-specifi c working hours

of these groups were weighted using the employment structure of

the euro area as a whole.

32ECB


June 2008

After adjusting for the structure of employment,

the spread of weekly hours worked inside the

euro area decreases by one fourth, to 9 hours per

week. Nevertheless, working weeks still appear

to be longest in Greece and – by far – shortest in

the Netherlands (see Chart 6).54

3.4 IMMIGRATION 55

Both the quantity and the quality of labour

supply in the euro area are important in order to

maximise welfare and future potential growth.

Immigration provides one channel for the euro

area to increase its labour supply along both

dimensions. Furthermore, it may help alleviate

shortages of particular skills in the labour

market, improving the allocation of labour

resources. It may offset some of the negative

effects of demographic change and, since

immigrants tend to be more mobile than native

workers, it may also help the labour supply

adjust to economic shocks. Immigration to the

euro area has increased signifi cantly over recent

decades. Chart 7 presents the net fl ow of

migrants into the euro area and into the United

States since 1980. It shows that net migration to

the euro area has been higher than to the United

States since 1997.56 Furthermore, migration to

the euro area prior to the late 1990s was mainly

the result of guest-worker programmes (1950s

and 1960s), family-reunifi cation (1970s) and

asylum-seeking (late 1980s and early 1990s,

when political events and ethnic confl icts

increased). More recently, euro area immigration

appears to have entered a new phase, possibly

as a result of EU enlargement, globalisation, and

changing immigration policies in some

countries, with an increase in the number of

immigrants (also of different origin) looking for

employment.

In the case of Greece, the relatively long hours even after this 54

adjustment may refl ect in part the relatively high share of retail

and tourist businesses with long operating hours.

Prepared by M. Ward-Warmedinger.55

This measure therefore estimates the immigration of individuals 56

from outside each area, it does not consider cross-state or cross-

euro area migration.

Chart 7 Net migration to the euro area and the United States

(in millions of persons)

-0.5

0.0

0.5

1.0

1.5

2.0

-0.5

0.0

0.5

1.0

1.5

2.0

1980 1985 1990 1995 2000 2005

euro areaUS

Source: Eurostat. Note: For this chart only, net migration is the difference between immigration into and emigration from the respective area during the year, often measured as the difference between the total population on 1 January and 31 December for a given calendar year, minus the difference between births and deaths (or natural increase). To the extent that this data does not capture illegal immigration, it may underestimate immigrant fl ows.

Chart 6 Weekly working hours in 2007: corrected for the population structure effect 1)

(differences, in hours, with respect to the euro area average)

-8

-6

-4

-2

0

2

4

6

-8

-6

-4

-2

0

2

4

6

1 Belgium

2 Germany

3 Greece

4 Great Britain

5 Spain

6 France

7 Italy

8 Luxembourg

9 Netherlands

10 Austria

11 Portugal

12 Slovenia

13 Finland

14 Denmark

15 Sweden

16 United Kingdom

17 United State

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

observed

adjusted2)

Sources: EU-LFS (spring data) and NBB calculations. Note: 15 to 64 years old.1) For the United States, no detailed data are available to calculate an adjusted number of weekly working hours.2) Remaining difference, after correction for the population structure (composition of the population by gender, age, education level and professional status).

33ECB


June 2008

3 MAIN TRENDS IN

LABOUR SUPPLYWhilst the potential for signifi cant economic

gains from immigration exists, realising these

gains depends on the characteristics of

immigrants relative to nationals and immigrants’

successful integration into the labour market.

The future challenges facing the euro area

suggest the need for a great variety of skills. For

example, immigration may help meet the future

demand (see Chapter 5) for services such as

nursing, household care, childcare, health care

and eldercare arising from population ageing,

and for high skilled workers. Furthermore, it is

important that immigrants’ skills and

qualifi cations are effectively utilised.57

Table 7 shows that the increase in immigration

to the euro area since the mid-1990s was

composed largely of females, non-EU15

nationals and prime age individuals (25-54 age

group).58 Whilst nearly half of the existing

population of immigrants in 2007 were low-

skilled persons, the fl ow of new immigrants

since 1996 has mainly been made up of medium

and highly skilled labour. This may partly refl ect

selective immigration policies in place in many

EU Member States, which try to attract highly

skilled immigrants (discussed further in

Section 4.3). The increase in immigration from

1996 to 2007 was largest in Spain, Greece,

Portugal and Luxembourg (see Table 29 in

Annex 3). The extent to which migration fl ows

have contributed to labour supply also varies

strongly across euro area countries (see Table 8,

Table 29 in Annex 3 and also Box 2). One

particular form of worker immigration – namely

daily cross-border commuting – has increased

three-fold over the past 10 years (see the Box in

Work by the OECD (2006) “Gaining from migration: towards a 57

new mobility scheme” suggests that the over-qualifi cation rate is

two to three times higher for foreign born relative to native born

in some euro area countries.

In the absence of more detailed data on immigration, it is not 58

possible to provide details on the explanation of this trend.

However, one can speculate that family reunifi cation, an increase

in female labour market participation and the regularisation of

illegal work in some countries may have played a role in these

developments.

Table 7 The composition of the working age population by nationality

Average annual change (p.p)

Level (%)

1996-2001 2) 2002-2007 3) 2007

Total working age population 1)

Nationals -0.1 -0.2 91.9

Other EU15 citizens 0.0 -0.1 1.8

Non EU15 citizens 0.1 0.3 6.3

of 12 new

member states n.a. n.a.5) 0.9

of Non EU27 n.a. n.a.6) 5.3

Non-national working age population 4)

According to genderMales -0.4 0.1 -0.3 0.0 50.0 50.1Females 0.4 -0.1 0.3 0.0 50.0 49.9

According to age15-24 years old -0.7 -0.1 -0.2 -0.2 15.9 17.525-54 years old 0.3 0.2 0.5 0.0 73.5 64.355-64 years old 0.3 -0.1 -0.3 0.2 10.6 18.2

According to education levelLow -0.3 -0.2 -0.6 -0.1 48.6 36.6Medium -0.1 -0.2 0.9 0.2 34.4 42.1High 0.5 0.5 0.4 0.2 16.7 21.0

Participation ratesNationals 0.4 0.2 70.9

Other EU15

citizens 0.0 0.4 73.7


of 12 new


of Non EU27 n.a. n.a.6) 68.3

Employment ratesNationals 0.8 0.3 65.9

Other EU15

citizens 0.5 0.2 67.6


of 12 new


of Non EU27 n.a. n.a.6) 57.7

Source: EU-LFS and ECB calculations. Note: 15 to 64 years old. To the extent that this data does not capture illegal immigration, it may underestimate the stocks and fl ows of immigrants.1) The non-national population is separated into non-national EU15 citizens and non-national non-EU15 citizens. For the period 2005-07, this last group is further split into the 12 new member states (which together with the EU15 form the EU27) and non-national non-EU27 citizens. 2) It is important to note that data for Ireland start in 1998, data for Portugal and the Netherlands in 1999, and data for Slovenia in 2002. 3) Irish data only available until 2004, Italian data only available for 2005-07. 4) The numbers in italics show the respective values for the national working age population. 5) The average annual change 2006-07 for total working age population is 0.1, for participation rates 0.6 and for employment rates 1.4. 6) Average annual change 2006-07 for total working age population is 0.2, for participation rates 0.6 and for employment rates 1.3.

34ECB


June 2008

Annex 4). Nevertheless, while the employment

level of migrants from other EU countries is

similar to or even higher than that of nationals

for most euro area countries, overall participation

and employment rates for non-EU nationals

lagged behind those of nationals in 2007,

especially in Belgium, Germany, France, the

Netherlands and Finland. There is some

evidence in some countries of decreasing

employment rates for non-nationals from 2002

to 2007. On the other hand, in Spain, Greece,

Italy and Portugal, the employment rate for non-

EU nationals was even higher than for nationals

in 2007 (see Table 8).59

These differences in participation rates may refl ect a country’s 59

history of immigration. For example, countries where

immigration is a new phenomenon may experience a high

proportion of immigrants who are driven by the economic

motivation to enter the labour market and fi nd a job. Countries

with a longer history of immigration may receive immigrants

driven by more varied and potentially less labour-market-driven

motivations, including, for example, family reunifi cation.

Table 8 Immigration in euro area countries

Country % of working age population 2007 Participation rates 2007 Employment rates 2007

NationalsOther EU15

Non-EU15

of which

12 NM

of which Non-

EU27 NationalsOther EU15

Non-EU15

of which

12 NM

of which Non-

EU27 NationalsOther EU15

Non-EU15

of which

12 NM

of which Non-

EU27

Belgium 90.9 5.3 3.8 0.5 3.2 67.0 69.1 55.7 71.1 53.2 62.4 62.4 40.6 62.1 37.1

Germany 89.6 2.8 7.6 1.0 6.6 76.7 76.7 63.8 73.3 62.4 70.6 69.6 51.5 63.7 49.7

Ireland n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a.

Spain 87.1 1.5 11.4 2.3 9.1 70.5 70.2 79.7 82.4 79.0 65.3 62.7 70.0 72.9 69.2

Greece 93.8 0.3 6.0 1.0 5.0 66.6 53.6 73.8 71.7 74.2 61.1 48.8 67.9 65.6 68.4

France 94.1 2.1 3.8 0.2 3.6 70.0 72.8 59.3 68.0 58.9 64.2 67.8 44.9 53.6 44.5

Italy 94.2 0.3 5.5 1.0 4.5 61.9 60.0 73.0 76.7 72.2 58.4 57.9 67.4 71.1 66.6

Luxembourg 57.9 37.7 4.4 1.1 3.3 61.7 71.6 64.9 63.4 65.4 59.4 69.3 57.7 61.4 56.4

Netherlands 95.7 1.6 2.7 0.2 2.5 79.1 78.0 57.6 72.3 56.3 76.7 75.8 51.8 67.7 50.4

Austria 88.7 2.2 9.1 1.6 7.6 74.2 77.2 67.8 75.1 66.3 71.3 73.1 59.5 69.6 57.3

Portugal 96.3 0.4 3.3 0.2 3.1 73.4 74.6 82.8 72.9 83.5 67.5 69.3 70.6 67.8 70.8

Slovenia 99.3 n.a. 0.7 n.a. 0.7 71.7 n.a. 65.7 n.a. 65.1 68.4 n.a. 59.7 n.a. 59.0 Finland 98.0 0.4 1.6 0.4 1.2 77.4 85.1 67.8 80.0 63.9 71.6 77.8 53.6 74.4 47.0

Euro area 91.9 1.8 6.3 0.9 5.3 70.9 73.7 69.6 77.1 68.3 65.9 67.6 59.3 68.7 57.7

Denmark 94.6 1.1 4.4 0.2 4.2 81.2 76.6 61.1 76.3 60.3 78.5 73.8 53.9 71.8 52.9

Sweden 95.0 2.1 2.9 0.3 2.6 80.4 75.4 65.1 74.1 63.9 75.1 70.8 52.3 56.0 51.9

United

Kingdom 92.2 1.8 6.0 1.4 4.6 75.2 76.9 71.0 84.7 66.7 71.4 71.5 65.2 79.6 60.7

Sources: Eurostat, LFS and ECB calculations. Notes: 15 to 64 years old. For Ireland: data available for 1998-2004 only; for Italy: for 2005-07 only; for the Netherlands and Portugal: data start in 1999; for Slovenia: data start in 2002; for Sweden: data start in 1997. For Greece, fi gures from the LFS differ signifi cantly from those of the 2001 Population Census. According to the 2001 Population Census non-EU and other EU15 citizens accounted for 7.7% and 0.5% of the working age population in Greece in that year. Numbers in italics are based on fi gures smaller than the Eurostat’s reliability limits.

35ECB


June 2008

3 MAIN TRENDS IN

LABOUR SUPPLY

Characteristics of immigrants for selected countries (2007)

Ireland Spain Italy Austria Immigrants Natives Immigrants Natives Immigrants Natives Immigrants Natives

Quantity

% Total Population 6.6 12.9 5.8 11.3

Gender

% Male 50.6 50.2 49.9 50.7 49.0 50.1 49.5 49.8

Age

15-24 20.2 23.5 18.1 16.3 14.7 15.6 17.0 18.0

25-54 73.5 62.1 76.9 66.6 81.1 65.4 72.0 64.7

55-64 6.3 14.4 5.1 17.0 4.2 19.0 11.0 17.3

Education

Low 20.8 39.1 46.3 50.3 51.8 48.9 38.4 23.9

Medium 28.5 36.7 33.8 21.2 37.4 39.0 45.7 61.9

High 38.3 23.3 19.1 27.0 10.9 12.1 15.8 14.2

Participation Rate 66.6 68.8 78.5 70.5 72.4 61.9 69.6 74.2

Unemployment Rate 6.3 4.4 12.0 7.3 7.6 5.7 10.8 3.9

Source: EU-LFS.Note: 15-64 year old.* Data for Ireland refer to 2004.

Box 2

RECENT EXPERIENCES WITH IMMIGRATION IN EURO AREA COUNTRIES 1

Although most European countries have experienced an increase in immigration during the

last 20 years, country experiences with immigration have differed (in terms of the magnitude of

infl ows, the characteristics of immigrants and the impact of immigration on the native population).

While some countries have a long history of immigration (e.g. Germany, France and Austria),

there are several countries for which large-scale immigration is a relatively new phenomenon (e.g.

Ireland, Spain or Italy). Numerous factors account for these cross-country differences, including

historical ties with a host country, a common language, geographical proximity and the extent of

labour market opportunities, which can affect both the magnitude and duration of immigration to

a particular country. This box focuses on the immigration experiences of Spain, Italy, Ireland and

Austria, which have attracted the majority of permanent immigrants in recent years.

Immigration in Spain and Italy

The immigration experiences of Spain and Italy have been fairly similar in recent years. In both

countries immigrants have become an important and a fast growing share of the population.

Spain has faced the highest net migration rates in Europe, with an average of 700,000 new

immigrants per annum during the last two years. This consistently high infl ow pushed up the

percentage of non-nationals in the Spanish population from 2% in 2000 to 13% seven years later

(see Table). Italy also experienced consistently high infl ows of foreign workers, although at a

relatively lower rate, increasing the share of non-nationals from 2.7% at the end of 2002 to 6% of

the total population in 2007.

1 Prepared by A. Rosolia, A. Lacuesta, Y. McCarthy and A. Stiglbauer.

36ECB


June 2008

Geographical proximity and historical ties are the main factors underlying migration to Spain and

Italy. Immigrants from South America account for over one third of the immigrant population

in Spain, and immigrants from Africa, especially from Morocco, accounted for 20% of the total

stock in 2006. Albania, Morocco and Romania are the main source countries of migration to Italy,

accounting for one-third of all immigrants in 2006. Despite the retention of some restrictions on

the free movement of workers from the 2004 EU new Member States, the fl ow and the share of

immigrants coming from these countries have signifi cantly increased. For example, in Spain,

those origin countries represented 2% of all immigrants in 2000, compared with more than 15%

just seven years later.

Regarding individuals’ characteristics, in both countries, non-nationals are much younger than

nationals. While about 65% of natives are concentrated in the 25-54 age group, this number

increases to above 76% for the non-national population. There are no important educational

attainment disparities between non-nationals and nationals in either country.

Despite the magnitude of immigrant fl ows and the similarity of educational attainment of

immigrants and natives, there is no evidence that migration has reduced the job opportunities

for residents in either country. Indeed, the simple correlation across regions between the

participation and employment rates of Italian citizens aged 25-54 and the share of foreigners in

the same population age-group is not statistically signifi cant.2 Research conducted at the Bank

of Italy relates natives’ labour market outcomes to the presence of foreign citizens in a province

over the decade 1993-2003, taking into account a host of individual characteristics as well as the

characteristics of local labour markets that could affect simultaneously Italian citizens’ labour

market outcomes and the overall presence of foreigners. The results show that male and female

employment and participation display a positive correlation with the presence of foreigners of

the same sex. Carrasco et al (2008) undertake an analysis in this regard for the Spanish case,

fi nding that foreigners have no signifi cant negative impact on natives’ occupational opportunities.

Rather, the increase in foreign workers has occurred at the same time as the increase in female

participation, and some research has linked the two phenomena. This evidence is confi rmed by

independent work of The Economic Bureau of the President in Spain, which provides evidence

that immigration facilitates female participation in the labour market by increasing the supply of

domestic help, thus easing the combination of work and family life for women.

One possible explanation for the lack of a negative relationship between immigrants’ and natives’

labour market outcomes is the fact that highly skilled immigrants have been willing to take jobs

requiring less skill than their educational attainment. Fernández and Ortega (2006) show that

this matching problem is more important for immigrants than for natives (see also Chapter 5).

Moreover, immigrants’ contractual conditions have been more fl exible than those of natives with

the same characteristics. In Spain, for instance, the share of immigrants on a temporary contract

is around 60%, compared with 30% for natives. Unemployment has also been much higher for

non-nationals than for nationals (respectively 12%/8% for non-nationals and 7%/6% for nationals

in Spain/Italy). However, at least part of this difference can be attributed to immigrants’ lack

of experience in their destination country, and over time the gap with the native population is

expected to close (Amuedo-Dorantes and de la Rica, 2007; Fernández and Ortega, 2006).

2 Because immigrants tend to locate where job opportunities are richer, such simple correlations might hide some crowding-out of

comparable native workers.

37ECB


June 2008

3 MAIN TRENDS IN

LABOUR SUPPLYImmigration in Austria

At 11% in 2007, Austria has a large non-national population. At the end of the 1960s, the

immigrant population share was only 2%. Thereafter, it increased gradually to 4% at the end

of the 1980s, then, in the wake of the fall of the “iron curtain” and the Yugoslav wars, a large

number of immigrants entered Austria within just a couple of years. By 1995, the immigrant share

had risen to more than 8%. Since then, it has risen continuously, although at a slower pace. Most

immigrants stay permanently, but seasonal work is also signifi cant. The share of immigrants in

the workforce is more than 12%, with the largest groups of non-national workers coming from

the former Yugoslavia, Germany, Turkey, Hungary, the Czech Republic and Slovakia, Poland

and Romania.

Non-nationals in Austria are on average younger than nationals: 11% of the non-national

population is between ages 55 and 64, compared with 17% of natives. Non-nationals tend to

be less educated than nationals, with 38% of the foreign population holding only a primary

education, compared with 24% of nationals. This partly underlies the fact that immigrants work

disproportionally in industry, in the construction sector, in tourism and agriculture. Seasonal

work and commuting are very common in the latter sectors. Illegal infl ows of immigrant workers

are most likely substantial, particularly in the household and personal service sector, which

includes cleaning and nursing services.

The available empirical evidence suggests that the aggregate effects of immigration on native

workers’ unemployment and wage growth are small or insignifi cant. Econometric studies are

mostly from the mid-1990s and exploit the then-sudden increase of foreign workers. They fi nd

that increased immigration had no negative employment or wage effects for native women. For

almost all groups of men, these studies fi nd only a slight deterioration of employment prospects.

Low-income men also faced lower wage growth, whereas high-income males appeared to

experience a wage gain associated with the increase in immigration.

Austria faces several challenges with regard to the integration of immigrants. For example, the

unemployment rate for non-nationals in 2007 stood at about 11% (whereas that of nationals was

merely 4% − see also Chapter 5). The difference between the performance of immigrant children

and nationals in the PISA studies was also one of the largest observed over all participating

countries (OECD, 2007b). Without appropriate countermeasures, Austria risks perpetuating the

lower labour market chances of immigrants and their descendants.

Immigration in Ireland

Ireland’s experiences with migration changed dramatically during the 1990s as Ireland rapidly

moved from being a country of net emigration to one of signifi cant positive net infl ows. Since

1996, net migration to Ireland has been positive, and the most recent numbers show a net

migration of 71,800 individuals for the twelve months to April 2006 (1.7% of the population).

As a result, there has been a rise in the non-national proportion of the population, from about

6% in 2002 to about 10% in 2006. The change in Ireland’s migration experience has also had

an important impact on labour supply growth. In the period 2000 to 2005, labour supply grew

by almost 3%, with migration accounting for just under half of this growth. By comparison, in

the early 1990s migration subtracted from labour supply growth. This change in the quantity

of immigrants in Ireland was accompanied by an important change in the origin country of

38ECB


June 2008

3.5 THE SUPPLY OF KNOWLEDGE AND SKILLS 60

Labour quality can be improved through

investment in human capital. Human capital

consists of the ability, skills and knowledge

embodied in the general population that are

accumulated through schooling, training and

experience.61 This section begins with an

examination of a traditional measure of human

capital − that is, educational attainment − as

captured by the highest level of education

attained by individuals in the euro area. It then

examines alternative measures to account for

the quality of educational attainment. Box 4

assesses the growth in labour quality in the euro

area over time.

3.5.1 EDUCATIONAL ATTAINMENT

Improving the stock of human capital has

been formally identifi ed through the Lisbon

strategy as a key area of potential growth

within the euro area. The level of educational

attainment within a country serves as one

potential indicator of the stock of such capital.

Table 9 shows educational attainment of euro

area countries as captured by the proportion of

the adult population that has received various

levels of education over time.

The average proportion of the 30-54 year

old population with a high level of education

(tertiary education) in the euro area was 16.1%

in 1992. By 1999 this had increased to 20.1%,

and the fi gure stood at 24.6% in 2007. On the

other hand, there has been a fall in the proportion

of the population with only a low level of

education, and a subsequent rise in those with

a medium level. However, the rate of increase

in this measure of the stock of human capital

differs across euro area countries. For example,

Ireland registers the largest increase in persons

with a high level of education between 1999

Prepared by P. Cipollone, Y. McCarthy and K. McQuinn. This 60

contribution has benefi ted from the comments of P. Montanaro,

Bank of Italy.

In much of the initial empirical work addressing this issue, 61

human capital has been measured by educational attainment.

Educational attainment, in this regard, is taken to be the number

of years of schooling received by an individual.

immigrant infl ows. In the late 1980s, over half of immigrants entering Ireland came from the

United Kingdom. However, over time this proportion fell signifi cantly, reaching about 17% in

2006. In recent years, immigrants from the new EU Member States have been the most important

contributors to immigrant infl ows in Ireland.

Focussing on characteristics, Labour Force Survey data available to 2004 (see Table) show that

immigrants in Ireland are young relative to the native population. Data for 2004 indicate that

73.5% of the non-national population is in the prime working age category, 25 to 54 years old,

compared with 62.1% of natives. Immigrants in Ireland also tend to be more highly educated than

natives, as over 38% of non-nationals hold tertiary education compared to 23.3% of nationals.

Despite this higher educational attainment, non-nationals face higher unemployment rates relative

to nationals, at 6.3% and 4.4% respectively in 2004. 3 A study conducted by Barrett et al (2006)

using data from a national household survey for 2003 found that immigrants in Ireland tended to

hold occupations of a lower level compared with those held by natives, once characteristics such

as age and education had been taken into account. Barrett and McCarthy (2007) examined the

earnings of immigrants relative to natives in 2004, fi nding that immigrants earned 18% less than

natives on average, controlling for education and years of work experience. This gap was more

pronounced for immigrants from non-English speaking countries and for those immigrants with

a third-level qualifi cation.

3 However, this fi nding of a higher rate of unemployment for immigrants is not unusual, as immigrants could possess lower levels of

location-specifi c human capital when they arrive in a host country. Over time it would be hoped that immigrants would assimilate into

the labour market and that this rate of unemployment would fall.

39ECB


June 2008

3 MAIN TRENDS IN

LABOUR SUPPLY

and 2007 at 12.8 percentage points, followed

by Spain with a 9 percentage point increase.

Furthermore, in 2007, the proportion of the

population with a high level of education was

highest in Finland, Belgium and Ireland, while

it was lowest in Italy, Portugal and Austria.

An examination of educational attainment by

gender shows that both males and females in the

euro area registered an increase in educational

attainment between 1992 and 2007, and by 2007

there was very little difference between male

and female educational levels in the euro area.

Table 9 Educational Attainment of 30-54-Year-Old Population

(% of population by highest level of education attained)

1992 1999 2007Country Low Medium High Low Medium High Low Medium High

Belgium 47.1 30.4 22.5 40.4 31.8 27.7 29.4 37.3 33.3

Germany 17.4 58.8 23.8 17.5 57.4 25.1 14.1 60.1 25.8

Ireland 56.9 25.8 17.3 44.1 36.0 19.9 30.6 36.7 32.8

Greece 60.5 26.0 13.5 46.2 35.0 18.8 36.4 40.1 23.5

Spain 76.1 10.9 12.9 62.2 16.4 21.4 46.6 23.0 30.5

France n.a. n.a. n.a. 37.1 42.1 20.8 29.4 43.4 27.2

Italy 64.8 27.2 8.1 53.9 35.4 10.7 45.3 40.8 13.9

Luxembourg 64.1 22.8 13.1 35.0 45.5 19.5 32.7 38.9 28.4

Netherlands n.a. n.a. n.a. 33.8 42.6 23.6 24.5 44.1 31.4

Austria n.a. n.a. n.a. 22.6 62.3 15.1 18.5 63.1 18.5

Portugal 79.3 9.0 11.7 80.9 9.9 9.1 72.2 13.9 13.9

Slovenia n.a. n.a. n.a. 24.3 60.0 15.7 17.1 59.6 23.3

Finland n.a. n.a. n.a. 23.7 42.8 33.5 15.4 44.2 40.4

Euro area 47.4 36.5 16.1 39.3 40.6 20.1 31.9 43.5 24.6

Denmark 23.3 53.7 23.0 19.5 51.8 28.7 23.1 43.2 33.7

Sweden n.a. n.a. n.a. 20.5 49.2 30.3 12.1 55.8 32.0

United Kingdom 49.4 31.0 19.6 36.2 36.0 27.8 26.6 40.7 32.6

Sources: Eurostat, LFS and ECB calculations.Notes: 30 to 54 years old. Data include vocational training to the extent that a qualifi cation has been gained. See Annex 2 for details on this and defi nitions of low, medium and high. See Table 30 in Annex 3 for this information for prime age individuals only. See Table 31 for information on education attainment by cohort.

Table 10 Educational Attainment: Subject specialisation

(2006)

SubjectGeneral

Programmes

TeacherTraining

and Education

HumanitiesLanguages

and Art

Social Science Businessand Law

Science Math

Computing

Engineeringand

Manufacturing Agriculture

Health and

Welfare Services

Belgium 6.6 9.1 6.2 22.2 6.8 28.4 1.6 12.7 6.4

Germany 6.0 5.2 3.8 28.3 2.3 34.2 3.0 8.7 8.4

Ireland 50.8 3.9 4.0 11.1 8.8 8.1 1.7 6.4 5.2

Greece 47.6 3.0 5.9 13.1 4.1 13.7 1.4 6.0 5.2

Spain 30.2 6.1 5.8 22.8 6.1 15.6 1.3 8.6 3.5

France 1.6 0.9 10.2 32.2 9.3 29.3 3.7 8.0 4.8

Italy 0.1 10.2 13.0 31.5 11.8 22.4 2.2 4.8 3.9

Luxembourg 8.3 5.7 9.3 29.7 7.0 15.0 1.8 7.2 16.1

Netherlands 8.3 9.4 4.9 27.2 3.5 19.2 2.9 15.6 8.9

Austria 8.3 4.1 3.8 26.6 1.3 34.6 4.5 5.5 11.2

Portugal 3.3 8.8 20.9 25.7 19.0 12.7 0.9 6.0 2.7

Slovenia 8.5 4.7 2.1 26.3 1.2 37.9 3.2 5.5 10.6

Finland 13.4 3.2 4.4 17.6 2.3 29.3 4.9 12.8 11.9

Euro area 15.0 5.9 7.5 24.7 6.8 22.6 2.4 7.9 7.3

Denmark 1.8 4.8 7.2 30.3 5.0 26.7 3.1 15.1 6.0

Sweden 11.3 8.3 4.9 20.7 2.9 26.6 2.3 15.6 7.5

United

Kingdom 0.7 6.6 11.7 25.5 11.0 21.5 1.7 14.0 7.4

Sources: LFS and ECB calculations. Ireland refers to 2005 data.

40ECB


June 2008

For example, in 1992, respectively 19.4/12.8%

of males/females aged 30-54 years had a high

level of education in the euro area. These fi gures

increased to 21.8/18.4% in 1999 and 24.7/24.5%

in 2007.

Table 10 provides information on the type of

skills acquired through education in euro area

countries. More specifi cally, the table shows

educational attainment across euro area countries

in 2006 (the latest year available) according to

the main subject studied. The results show that

“engineering” (manufacturing and construction)

and “social science, business and law” are

generally among the top two most studied

disciplines in the majority of euro area countries.

In Ireland, Greece and Spain, however, “general

programmes” is the most popular area of study.62

In general, there has been little change in the

respective rankings of subject take-up across euro

area countries since 2003, the earliest year for

which data are available.

3.5.2 THE QUALITY OF EDUCATIONAL

ATTAINMENT

While cross-country comparisons of the

quantity of education received are important and

informative, attention has increasingly focussed

more on differences in the quality of education

received across countries. In attempting to

measure the quality of educational attainment, a

growing line of research uses indexes based on

scores obtained by students on cognitive tests.

Cross-country information on such measures

is provided under a number of different

international auspices. The IEA-TIMSS (Trends

in International Mathematics and Science Study)

collects educational achievement data at the

fourth and eighth grades in mathematics and

science. Comparable data are available for 1995,

1999 and 2003.63 ,64 The IEA-PIRLS (Progress

in International Reading Literacy Study)

examines literacy progress at the fourth grade

(9-and 10-year-olds). The fi rst survey was

conducted in 2001 on about 150,000 students

across 35 countries (including fi ve euro area

countries). A second survey was conducted in

2006. The OECD has launched the Programme

for International Student Assessment (PISA),

an internationally standardised assessment of

the competences of 15-year-old students. The

assessment covers the domains of reading,

mathematical and scientifi c literacy, and includes

all euro area countries with the exception of

Slovenia. The fi rst survey, conducted in 2000,

focused mostly on reading literacy; the second

was conducted in 2003 with a major focus on

math; science literacy is the main focus of the

third survey, conducted in 2006. The development

of these databases is of considerable interest,

particularly in the context of studies examining

cross-country income differentials.

“General programmes” refers to all other categories of education. 62

It is important to note that these categories may be somewhat

sensitive to a country’s educational system. For example, many

students at the degree level in Ireland would take a general ARTs

degree, included under the “General Programmes” defi nition.

However, in another country degree programmes may be more

specialised, allowing them to be categorised under “Social

Science, Business and Law” or “Humanities, Languages and

Art”.

The IEA (International Association for the Evaluation of 63

Educational Achievement) is an independent, non-profi t,

international cooperative of national research institutions and

governmental research agencies, established in 1959.

For more on this, along with a summary of results from both 64

PIRLS and TIMSS, see Montanaro (2007).

Box 3

HUMAN CAPITAL AND GROWTH 1

Initial growth studies incorporating the role of human capital such as Barro (1991), Benhabib

and Spiegel (1994), Barro and Sala-i-Martin (1995) and Sala-i-Martin (1997) had focussed

solely on the role played by educational attainment in terms of the number of years of schooling

1 Prepared by P. Cipollone, Y. McCarthy and K. McQuinn. This contribution has benefi ted from the comments of P. Montanaro,

Bank of Italy.

41ECB


June 2008

3 MAIN TRENDS IN

LABOUR SUPPLY

Table 11 compares educational attainment of

euro area countries as captured by the traditional

quantity measure, highest level of education

attained, with the PISA results for 2006 − the

latest year for which data are available. While

the quality and quantity of education are two

different concepts, the results in the table

nonetheless demonstrate a positive correlation

between the rankings of countries in terms of

the proportion of the population with a high

level of education and the rankings based on the

PISA profi ciency scores. For example, Finland

registers the highest proportion of its population

with a high level of educational attainment

in 2006, at 38.7%, followed by Belgium, at

32.6%. Finland also ranks the highest across the

three PISA categories, while Belgium and the

Netherlands are typically among the top four

countries when ranked by both the PISA average

scores and the proportion of the population with

a high level of educational attainment. In this

context, Greece would appear to be an outlier.

It registers among the lowest of the euro area

countries in terms of the average scores across

the three PISA categories, while it enjoys an

average level of high educational attainment

relative to the other euro area countries.

A key message emerging from this section is

that euro area countries have generally been

successful in increasing their stock of human

capital over time. However, there is scope for

further improvement, particularly since some

euro area countries still lag far behind the

average human capital level in the euro area,

as well as the level of human capital in other

advanced economies. Moving beyond the

human capital level of the euro area population,

Box 4 examines developments in the quality

of the euro area workforce between 1992

and 2005. This analysis suggests that labour

quality growth is likely to decline over the next

decade. This slowing of labour quality growth

could have important ramifi cations for labour

productivity growth and potential output. In

an attempt to ensure continued labour quality

growth commensurate with past rates, policy

should be directed towards promoting further

increases in educational attainment and on-the-

job training (see also Section 4.4 and Box 8).

in a country. These studies generally found that educational attainment was positively correlated

with the growth rate of GDP per capita across countries. The provision of databases such as

Trends in International Mathematics and Science Study (TIMSS) and Progress in International

Reading Literacy Study (PIRLS), however, enables studies of income differentials to focus on

the role played by enhanced educational quality. For example, TIMSS formed the basis for the

measurement of labour-force quality in studies by Hanushek and Kimko (2000) and in Barro

(2001). Both of these studies used the TIMSS database as an enhanced indicator of educational

attainment in growth regressions. The Hanushek and Kimko (2000) results are particularly

interesting as they emphasize how the explanation of cross-country growth is affected by the

inclusion of quality measures. Their estimates suggest that a one-standard-deviation improvement

in a test performance (equivalent to 47 score points in PISA 2000 mathematics) could increase

the annual average growth rate of a country by 1 percentage point. 2 Recently, Hanushek and

Wößmann (2007) have extended the Hanushek and Kimko results to 50 countries (from 31),

exploiting all the available information gathered by IEA’s and OCSE PISA survey on test score

in math and science. The study supports the Hanushek and Kimko fi nding of a positive link

between educational quality and GDP growth. Additional work by Bosworth and Collins (2003),

Ciccone and Papaioannou (2005), Coulombe et al. (2004) and Coulombe and Tremblay (2006)

all fi nd that improvements in educational quality strongly outweigh increases in educational

quantity in infl uencing economic growth.

2 See Montanaro (2007) for a more detailed discussion of this issue.

42ECB


June 2008

Box 4

LABOUR QUALITY GROWTH IN THE EURO AREA AND EURO AREA COUNTRIES 1

Increases in the supply of highly educated and more experienced workers have the potential

to contribute positively to economic growth. Standard measures of labour input, such as total

hours worked, ignore these changes in the composition of the workforce, typically leading to an

underestimation of the contribution of labour to growth (see OECD, 2001). This box presents

estimates of the trend in labour quality growth for the euro area and euro area countries. The

estimates of labour quality are constructed in two steps. In a fi rst step, microdata are used to

derive weights for a number of worker groups with different characteristics. These weights refl ect

differences in productivity (measured by estimated relative wages 2) across workers groups, e.g.

those with university level education or more are on average more productive than those with

primary education and are thus given a larger weight. In a second step, these weights are used

to adjust data on total hours worked by worker-country groups to arrive at an index of labour-

quality-adjusted labour input. Labour quality growth is estimated as the difference between

1 Prepared by J. Turunen.

2 More specifi cally, time-varying weights are derived from predicted wages from cross-section regressions of individuals’ wages on their

human capital characteristics such as education and labour market experience (as proxied by age).

Table 11 Educational Attainment of Population and PISA scores, 2006

(% of total population by highest level of education attained; Pisa score)

Country 2006 PISA 2006 PISA 2006Low

EducationMedium

EducationHigh

EducationProfi ciency in Reading

Profi ciency in Maths

Profi ciency in Science

Percentage 30-54 year old population Average Score Total

Belgium 30.9 36.4 32.6 501 520 510 1531

Germany 15.0 59.3 25.7 495 504 516 1515

Ireland 31.7 37.1 31.2 517 501 508 1526

Greece 37.2 39.3 23.5 460 459 473 1392

Spain 48.3 22.3 29.4 461 480 488 1429

France 31.1 42.9 26.0 488 496 495 1479

Italy 46.1 40.5 13.4 469 462 475 1406

Luxembourg 34.0 41.6 24.4 479 490 486 1455

Netherlands 25.3 43.7 31.0 507 531 525 1563

Austria 18.4 63.0 18.7 490 505 511 1506

Portugal 72.1 14.2 13.7 472 466 474 1412

Slovenia 17.7 60.4 22.0 494 504 519 1517

Finland 15.6 45.8 38.7 547 548 563 1658

Euro area 32.8 43.1 24.1 491 497 503 1491

Denmark 17.1 46.3 36.6 494 513 496 1503

Sweden 12.6 56.3 31.1 507 502 503 1512

United Kingdom 27.2 41.7 31.1 495 495 515 1505

Sources: Educational attainment data are from Eurostat, LFS and ECB calculations. PISA data are from the OECD.

43ECB


June 2008

3 MAIN TRENDS IN

LABOUR SUPPLY

quality-adjusted and raw total hours worked. Estimates of labour quality growth are based on a

number of assumptions and data sources and should thus be interpreted with great caution.3

Focussing on the period 1992-2005, estimates in Schwerdt and Turunen (2007) suggest that

euro area labour quality has grown on average by 0.48% year-on-year. Relatively strong labour

quality growth in most of the 1990s, driven by an increase in the share of those with tertiary

education and workers in prime age (35-54 years of age), was followed by lower labour quality

growth towards the end of the 1990s, possibly refl ecting the impact of robust employment

growth resulting in the entry of workers with lower skills. The euro area estimate of labour

quality growth refl ects substantial diversity across individual countries for which reliable

estimates are available, with estimates ranging from the lowest in Finland (0.18%) to the highest

in Spain (0.84%) (see Chart A). In line with other studies (see, e.g. Jorgenson, 2005, for the

G7 countries), the rise in the average level of educational attainment is the main driver of the

increase in labour quality over time, with a consistent relative contribution to labour quality

growth also across euro area countries. Chart A shows that overall labour quality growth has

signifi cantly benefi ted from increasing shares of highly educated employees in countries like

Spain, Ireland and Austria. Other countries, such as Finland and Germany, do not show such

signifi cant increases in the share of highly educated employees and, consequently, have lower

3 In particular, estimates of labour quality growth are based on the key assumption that relative marginal products of worker types are

refl ected in their relative wage rates. Various characteristics of labour markets, such as discrimination, union bargaining, signalling and

mismatch, may result in violations of this assumption. However, due to a lack of more direct measures, wages remain the best available

proxy of worker productivity. Furthermore, individual labour market experience is not directly observable in available household data.

Therefore, as is standard in the literature, age is used to proxy labour market experience. The weights are derived separately for men and

women, allowing e.g. for the age-earnings profi les to differ across gender. Levenson and Zoghi (2007) construct labour quality growth

estimates for the US based on birth-imputed experience measures and age and fi nd that using age results in a slightly lower estimates of

labour quality growth. Schwerdt and Turunen (2007) use detailed information from the European Community Household Panel (ECHP)

and the LFS on total hours worked and wages by worker groups along four dimensions − age (6 groups), education (3), gender (2) and

(for the euro area estimates) country (12) − to construct estimates of labour quality growth. For a more detailed description of the data

and methodology, see Schwerdt and Turunen (2007). Because of the reclassifi cation of education categories in the LFS that occurred in

the late 1990s and other breaks, estimates for country-years in which at least a 5% jump in an underlying share of total hours worked

within a single education category is observed (1998 in Ireland, 1998 in Finland, 1999 in Austria and 2004 in Greece) are excluded

from the calculation of time period averages shown in this box.

Chart A Labour quality growth and the contribution of education

(average annual growth rate in 1992-2005; percent)

Germany

Spain

Belgium

Austria

Ireland

Euro Area

FranceItalyPortugal

GreeceNetherlands

Finland0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

x-axis: growth in labour quality

y-axis: 1st order contribution of education

Source: ECB calculations based on estimates in Schwerdt and Turunen (2007).

Chart B Labour quality growth over time in the euro area

(annual growth rate; percent; 1992 to 2005)

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1992 1994 1996 1998 2000 2002 2004 2006

Source: ECB calculations based on estimates in Schwerdt and Turunen (2007).

44ECB


June 2008

fi rst-order contributions from education to labour quality growth over this time period. Overall,

these country results are broadly consistent with estimates from other studies.4

Chart B shows the growth in labour quality in the euro area since 1992. Looking forward, owing

to the ageing of the euro area population, the relative share of (the most productive) workers

of prime-age is likely to decline, putting downward pressure on growth in euro area labour

quality in the coming 10-15 years. This effect poses an additional challenge for sustaining labour

productivity growth in the euro area.

4 See Jorgenson (2005) for estimates for Germany, France and Italy; Card and Freeman, (2004) for Germany; Melka and Nayman (2004)

for France; Brandolini and Cipollone (2001) for Italy; Inklaar et al. (2005) for the EU4 (Germany, France, the Netherlands and the

United Kingdom) and EU-KLEMS for a number of European countries (see www.euklems.net).

45ECB


June 2008

4 STRUCTURAL

POLICIES AND

THEIR EFFECT ON

LABOUR SUPPLY

4 STRUCTURAL POLICIES AND THEIR EFFECT

ON LABOUR SUPPLY 65

An individual’s labour supply is characterised by

two key decisions: fi rst, whether to participate

in the labour market; and second, once in the

labour market, how many hours to work. Such

decisions are determined by preferences and

budget constraints over the life-cycle and are

greatly affected by institutions and structural

conditions.66 Empirically, a large literature has

documented the signifi cant impact of labour

and product market institutions on labour

supply across countries.67 This literature has

emphasized that the impact of institutions differs

across labour types.68 Institutional changes are

found to have the strongest impact on the labour

supply of individuals with a weaker attachment

to the labour market, such as younger and

older workers, females and (although less

evidence is available) immigrants.69 Against the

background of population ageing, globalization

and technological change, policies which

succeed in attracting, integrating and retaining

a high number of people in productive jobs and

increasing human capital cannot be understated

in their importance for increasing income levels

and for the sustainability of public fi nances in

the Member States.

This chapter thus takes a closer look at

the contributions of structural policies or

institutions to increasing euro area labour

supply. While these institutions will mainly be

considered separately below, the interaction of

different institutions (e.g. active labour market

policies and unemployment benefi ts) and the

dependency of their effect on macroeconomic

conditions is an important issue.70 Particular

consideration is given to institutions affecting

the subgroups in Chapter 3, namely: (i) Tax

and benefi t systems, and the design of pension

and early retirement systems, (ii) “Work/family

reconciliation” policies, including parental

leave, part-time work opportunities and child

care provision, (iii) Policies and institutions

affecting immigration and the integration of

immigrants into the labour market and society,

(iv) Educational and training policies.

This chapter’s main fi ndings include the fact

that reducing disincentives to work, such as

high marginal tax rates, high unemployment

benefi ts, early retirement schemes and weak

work availability requirements, can stimulate the

labour supply and employment of all workers, but

particularly those with a generally more tenuous

attachment to the labour market, such as women

and older workers. Female labour supply in

particular is supported by so called “work/family

reconciliation” policies. These policies help to

reconcile fertility and labour market participation

developments, with positive implications for

long-term labour supply and potential output.

From an economic perspective, the benefi ts

derived from immigration depend on both the

quantity and the characteristics of migrants

entering the euro area. The design of national

policies and institutions is important to set proper

incentives and to facilitate the integration of

immigrants into the labour market and society.

Prepared by A. Balleer, J. Turunen and M. Ward-Warmedinger.65

For example, the participation decision is affected by factors 66

infl uencing the attractiveness of entering the labour market,

relative to remaining outside it, such as the tax system and the

generosity and duration of unemployment benefi ts, institutions

that affect the incentives of workers (and fi rms) to invest in

education and skills, contractual arrangements and regulations

affecting the fl exibility of hours of work including childcare,

parental leave and/or part-time work opportunities. The second

decision is largely determined by the extent to which working

more hours results in higher current or expected net income. In

the case of tight regulation, the key decision may degenerate to a

decision about whether to participate full-time or part-time with

the precise number of hours in these categories fi xed by law or

social partners. This infl exibility may also lead to individuals

deciding not to enter the labour market at all. Additional aspects

of individual labour supply include joint household decisions

and decisions relating to the life-cycle (e.g. how many years

to participate and decisions relating to investment in human

capital).

While there exists a large number of studies that focus on 67

unemployment, employment and participation decisions have

not attracted as much attention in the literature. See for example,

Genre et al. (2005b) on participation rates in the European

Union, and Bassanini and Duval (2006) and Bertola et al.

(2002), who investigate employment and relative employment

of the sub-groups in OECD countries in 1982-2003 and

1960-1996 respectively.

See for example Antunes and Cavalcanti (2007), Genre et al. 68

(2007), Bassanini and Duval (2006), Bertola et al. (2002) and

Nickell and Layard (1999) for more details.

See for example Bertola et al, (2002), Amable et al. (2007), 69

Nickell and Layard (1999).

See Bassanini and Duval (2006), Carone and Salomäki 70

(2001), Bertola et al. (2002), Blanchard and Wolfers (2000),

Killingsworth and Heckman (1986).

46ECB


June 2008

Finally, educational systems and the amount

and effi ciency of national resources devoted

to education play a key role in determining

the labour supply of the young and innovation

capacity, overall labour quality within the euro

area and long-term wage levels. It is important

that national education systems are well funded

and effi cient, providing positive incentives for

young people, workers and fi rms to invest in

education and training, and for the effi ciency and

service orientation of educational institutions to

be improved. Furthermore, educational systems

have an important role to play in ensuring a

smooth transition from education to working life

and in providing the labour force with relevant

skills for the future. However, a general caveat

is needed: namely that policies involving higher

government spending must also be considered

from the perspective of the government budget.

The distorting effects of tax increases to fi nance

such policies also need to be taken into account.

4.1 TAX AND BENEFIT SYSTEMS 71

Tax and benefi t systems are a major explanatory

factor of the labour supply developments

observed in the euro area over the last decade.

Depending on their design, tax and benefi t

systems affect individuals’ incentives to engage

in paid employment in several ways. First, they

may affect the decision to enter paid employment.

High income support for persons not in

employment (e.g. unemployment benefi ts, social

assistance, benefi ts from disability schemes,

housing benefi ts) and high taxes on labour, such

as income taxes and social security contributions,

reduce the incentives for moving from inactivity

to activity, from informal (or activities in the

shadow economy) to formal work and from

unemployment to employment. Second, tax and

benefi t systems affect work effort or human capital

formation. In this respect, high labour taxes reduce

incentives to work longer hours, to move from

part-time to full-time work, to increase efforts and

learning to enhance future income prospects, and

to move to jobs with higher productivity. Third,

pension systems affect incentives to stay in the

labour force longer. Against this background, this

subsection describes through what channels the

features of euro area countries’ tax and benefi t

systems, as well as changes therein, may have

supported the rise in labour supply that was

identifi ed in the previous chapter.

4.1.1 UNEMPLOYMENT BENEFIT SYSTEMS

Unemployment benefi ts are intended to provide

fi nancial security in case of unemployment and

to increase matching in the labour market by

giving jobseekers suffi cient time to fi nd suitable

jobs that match their abilities. However, by

raising workers’ reservation wages (i.e. the

wage level below which jobs are rejected), high

unemployment benefi t levels and long benefi t

durations tend to reduce the unemployed

person’s job search intensity and willingness to

accept job offers, and are thus likely to result in

an increased incidence and duration of

unemployment (for a survey of the empirical

literature, see OECD, 2006b, Chapter 3).72

As indicated in Table 12, unemployment benefi t

levels differ widely across euro area countries.

Over the last two decades, unemployment

benefi ts, as measured by the OECD summary

measure of unemployment benefi t entitlements,

have tended to increase in the euro area. In

more recent years, however, according to

this indicator, several countries have reduced

income support during unemployment,

in particular Germany, France and the

Netherlands. At the same time, in the majority

of euro area countries, the net replacement

rate 73 of unemployed persons with below-

average incomes was raised between 2001 and

2005, while for persons with average incomes,

Prepared by N. Leiner-Killinger.71

Furthermore, Genre et al. (2007) fi nd that female labour market 72

participation declines with the size of unemployment benefi t.

Bertola et al. (2002) show that unemployment benefi ts increase

prime-age employment relative to employment of young workers.

Centeno and Novo (2007) distinguish between a substitution and

an income effect generated by unemployment benefi ts. They fi nd

that an extension of the entitlement period in Portugal highlights

the importance of the income effect on labour supply, which

mitigates the distortionary nature of unemployment benefi ts. The

results indicate, however, that the most constrained individuals

benefi t the least from the extension of the entitlement period.

The ratio of an individual’s (or a given populations’ (average)) 73

net-income when unemployed in a given time period and the

(average) net pre-unemployment income when employed in a

given time period.

47ECB


June 2008

4 STRUCTURAL

POLICIES AND

THEIR EFFECT ON

LABOUR SUPPLY

reductions and increases in the replacement

rate nearly balanced at the euro area level.

Over the past decade, the vast majority of

euro area countries tackled unemployment

benefi t administration by, for example,

tightening work availability conditions,

shortening the duration of benefi t payments

or tightening the eligibility conditions for

unemployment benefi t receipt (see Table 12).

In addition, several countries introduced in-

work benefi ts for workers working at low

wages (see OECD, 2006b, Chapter 3). On

average, these measures are found to have

supported employment creation by, inter

alia, raising the participation rate of persons

with low levels of education in particular

(see Chapter 3.1), and perhaps also by

supporting moderate wage increases, thereby

enhancing labour demand (see Box 5 for

reform experiences in Ireland and Germany).

4.1.2 LABOUR TAXES

Labour taxes, including income taxes and social

security contributions, affect labour supply,

equilibrium employment and hours worked, as

they drive a wedge between the marginal

productivity of labour and net income received.

In the presence of relatively high unemployment

insurance (i.e. net replacement rates), they may

have a particularly detrimental effect on

employment of workers with low incomes.74, 75

The effect of higher labour taxes (all else equal) on individual 74

labour supply is a priori ambiguous. On the one hand, higher

labour taxes may decrease disposable income, therefore

decreasing leisure and increasing the supply of labour (income

effect). On the other hand, labour supply may decrease, increasing

leisure, which is now cheaper (substitution effect). However, if

in general equilibrium, taxes are used to fund benefi ts such as

unemployment insurance. Negative income effects following

from higher labour taxes would tend to be cancelled out by the

positive income effect for benefi t recipients, while the substitution

effect goes in the same direction – reducing labour supply.

See e.g. Carone and Salomäki (2001).75

Table 12 Reforms of unemployment benefit systems in euro area countries

OECD summary measure of benefi t entitlements 1)

Net replacement rates 2) Reform policies 1994-2006 3)

67% of APW 100% of APW Shorter benefi t

duration

Tighter work availability conditions 4)

Tighter eligibility

conditions 5) level2005

change (p.p)1985-2005

change (p.p)2001-2005

level2005

change (p.p)2001-2005

level2005

change (p.p)2001-2005

Belgium 40.9 -2.2 2.4 71.0 2.0 56.0 1.0 + + +

Germany 24.2 -4.1 -5.2 78.0 -3.0 73.0 -2.0 + +

Ireland 33.7 5.4 4.0 70.0 3.0 59.0 5.0 +

Greece 13.3 6.1 0.2 65.0 2.0 47.0 3.0

Spain 36.0 1.6 -0.5 77.0 0.0 75.0 -1.0 - +

France 39.0 4.6 -4.5 83.0 0.0 67.0 -4.0 + +

Italy 32.5 32.1 -1.6 64.0 7.0 70.0 8.0 + +

Luxembourg 26.7 n.a. 0.0 90.0 1.0 89.0 0.0 +

Netherlands 35.3 -19.3 -17.5 86.0 1.0 70.0 0.0 + + +

Austria 31.9 2.5 0.3 72.0 -1.0 68.0 -1.0 + +

Portugal 39.5 17.9 -1.7 85.0 9.0 77.0 0.0 + + -

Slovenia n.a. n.a. n.a. n.a. n.a. n.a. n.a.

Finland 35.3 0.9 0.4 85.0 -2.0 76.0 -4.0 + +

Euro area 32.3 4.1 -2.0 77.2 1.6 68.9 0.4

Denmark 48.9 -4.2 -2.0 91.0 0.0 75.0 -1.0 + + +

Sweden 23.8 -4.1 0.2 89.0 -1.0 69.0 -3.0 + +

United

Kingdom 15.6 -5.1 -1.0 70.0 18.0 60.0 16.0 +

United States 13.5 -1.2 0.0 49.0 -2.0 56.0 -3.0

Sources: OECD (2006b) Employment Outlook, Chapter 3, OECD (2007b) Benefi ts and wages.Notes: Unweighted averages for the euro area. n.a. not available.1) The OECD summary measure of benefi t entitlements is supposed to measure the overall generosity of unemployment benefi t systems. It is defi ned as the average of the gross unemployment benefi t replacement rates for two earnings levels, three family situations and three durations of unemployment.2) Net replacement rates are defi ned after tax in the initial phase of unemployment but following any waiting period for a one-earner couple with two children aged 4 and 6. The pre-unemployment income situation relates to 67% (100%) of the average production worker wage (for further details, see www.oecd.org/els/social/workincentives).3) Evaluations are based on OECD (2006b) and NCB assessments. + (-) indicates an increase (decrease) in the respective indicator. Assessments for Denmark, Sweden, the United Kingdom and the United States up to 2004 only.4) Including tighter requirements for being available for work when offered a job. 2005 for Austria.5) Including tighter eligibility requirements for certain groups of persons, most often increases in the minimum period of insured employment required to receive unemployment benefi ts.

48ECB


June 2008

Moreover, the negative impact of social security

contributions on labour supply may be stronger,

the smaller the perceived link between workers’

social security contributions and benefi t

entitlements (‘perception effect’).76 Bassanini

and Duval (2006) and Silva (2005) document

that high tax wedges reduce employment. This

is supported by Nickell and Layard (1999).

Evidence also suggests that the impact of

taxation on labour supply may depend on the

way in which the public sector uses tax revenues

(see, e.g. Rogerson, 2007).

Table 13 surveys selected tax wedge

(the difference between employers’ cost of

employing workers and employees’ take-

home pay after tax) indicators for various

family situations. It shows that tax wedge

levels vary widely across euro area countries,

with differences amounting to more than

30 percentage points. In the majority of euro area

countries, tax wedges were reduced between

2001 and 2006, with the strongest reductions

taking place in Ireland (see also Box 5), which

has the lowest tax wedge levels across euro

area countries. The modest euro area-wide tax

wedge reductions between 2001 and 2006 tend

to underestimate long-term reform efforts, as

many countries undertook reforms lowering

labour taxes somewhat earlier.77 Overall,

there seems to be a tendency across euro area

See OECD (2007e), Chapter 4, pp. 170 for a discussion of the 76

literature.

For example, taking into account the year 2000 reduces the euro 77

area average tax wedges signifi cantly, i.e. by – 1.4 p.p for single

earners at 67% of the average worker wage, by – 1.2 p.p for a

one-earner married couple at 100% of the average worker wage

and by – 1.8 p.p for a two-earner married couple at 100% and

33% of the average worker wage.

Table 13 Changes in labour taxes in the euro area countries

Tax wedges 1) Unemployment trap 2) Low wage trap 2)

Single earner (67% of average worker wage)

One earner married couple (100%)

Two earner married couple (100%, 33%)

level 2006

change (p.p.) 2001-2006

level 2006

change (p.p.)2001-2006

level 2006

change (p.p.) 2001-2006

level 2005

change (p.p.)2001-2005

level 2006

change (p.p.)2001-2006

Belgium 49.1 -1.6 40.1 -2.5 41.0 -3.0 83.0 -3.0 58.0 2.0

Germany 47.4 -0.3 36.2 -0.6 41.5 -0.5 75.0 0.0 51.0 -2.0

Ireland 16.3 -1.1 2.3 -10.5 8.9 -7.9 76.0 3.0 53.0 7.0

Greece 35.4 0.3 41.5 1.8 40.0 1.5 59.0 3.0 19.0 1.0

Spain 35.9 0.6 33.6 0.9 35.4 0.2 80.0 0.0 26.0 2.0

France 44.5 -3.1 42.0 1.5 40.0 -0.6 81.0 0.0 35.0 -6.0

Italy 41.5 -1.2 35.1 -2.0 37.9 -2.5 71.0 12.0 33.0 4.0

Luxembourg 30.6 -0.6 13.0 -1.0 17.6 -0.7 88.0 0.0 51.0 8.0

Netherlands 40.6 1.7 37.0 8.8 36.8 5.6 86.0 7.0 70.0 5.0

Austria 43.5 0.6 36.9 2.0 37.7 1.9 67.0 0.0 37.0 2.0

Portugal 31.7 -0.5 26.6 -0.5 27.9 -0.3 81.0 0.0 20.0 -1.0

Slovenia n.a. n.a. n.a. n.a. n.a. n.a. 94.0 13.5 67.0 32.1

Finland 38.9 -2.5 38.0 -1.5 36.5 -1.8 76.0 -4.0 61.0 5.0

Euro area 38.0 -0.6 31.9 -0.3 33.4 -0.7 78.4 3.0 43.3 4.5

excl. SI 2.4 excl. SI 4.5

Denmark 39.3 -1.2 29.5 -1.1 34.4 -1.2 91.0 -1.0 82.0 -2.0

Sweden 46.0 -1.8 41.8 -1.1 41.7 -1.7 87.0 0.0 55.0 5.0

United

Kingdom 30.4 2.3 27.8 2.7 25.8 2.8 68.0 0.0 58.0 0.0

United States 26.4 -0.5 11.7 -3.1 19.3 -2.2 70.0 0.0 32.0 -2.0

Sources: OECD (2006d), “Taxing wages 2005-2006” and Eurostat Structural indicators database.Notes: Unweighted averages for the euro area. n.a. not available.1) The tax wedge captures income tax plus employee and employer social security contributions less cash benefi ts as a percentage of labour costs. It is displayed here for a single person without children with 67% of the average worker wage, for a one-earner married couple with two children aged 4 and 6 at 100% of the average worker wage and a two-earner married couple with two children, where one earner has 100% of the average worker wage and the other 33%. Ireland is based on the old OECD defi nition of earnings.2) The “unemployment trap” is defi ned as the percentage of gross earnings taxed away through higher taxes and social security contributions as well as benefi t withdrawal when an unemployed person takes up a job. It is measured here for a single person without children with 67% of the average earnings of a full-time production worker in the manufacturing industry.3) The “low wage trap” is defi ned as the percentage of gross earnings taxed away by higher taxes and reduced benefi ts when taking up a higher paid job. It is measured here for a single person without children, moving from 33% to 67% of the average earnings of a production worker.

49ECB


June 2008

4 STRUCTURAL

POLICIES AND

THEIR EFFECT ON

LABOUR SUPPLY

countries to focus reductions in tax wedges on

persons with low earnings and on two-earner

married couples, which may have supported the

increase in employment of workers with low

earnings and of women observed in the recent

past (see Section 3.2.1).

However, as indicated by the high levels and

sometimes even unfavourable developments

in the unemployment and low-wage-trap

indicators across euro area countries displayed

in Table 13, the fi nancial rewards for moving

from unemployment to employment and from

lower to higher earnings remained rather low

between 2001 and 2005.

High levels of labour taxation may also provide

incentives to engage in non-employment

activities such as home production. Bertola et

al. (2002), for example, document that the

employment differential between females and

males is positively affected by labour taxes, as

higher taxes lead women to engage more in

home production.78 In addition, high levels of

taxes on income and value added (see Table 14),

and high social security contributions may lead

to more labour being supplied in the shadow

economy and thus often in activities with low

productivity. In a differentiated Value Added

Tax (VAT) rate system, a lower VAT may prove

particularly supportive of labour supply in

sectors whose services are easily substituted for

do-it-yourself or work in the underground

economy.79 In addition, according to the

literature, such incentives for involvement in the

countries’ shadow economies further arise from

the desire to circumvent high legal labour

standards, such as, inter alia, certain safety

standards or high minimum wages, as well as

product market regulations such as, e.g. time-

consuming administrative procedures.80

Tax and benefi t systems also affect incentives to

build up human capital, e.g. through subsidies

paid for participation in education, as discussed

in Section 4.4.

See also Burda et al. (2006) and Freeman and Schettkat (2005). 78

In turn, Genre et al. (2005) do not fi nd a signifi cant effect of

labour taxes on participation.

For a discussion, see Copenhagen Economics (2007). Some 79

euro area countries have, over the past two decades, raised

their standard Value Added Tax (VAT) rate, although standard

rates and the levels of reduced rates differ signifi cantly across

countries (see Table 14).

See Schneider (2006) for a survey. In a narrow sense, the 80

shadow economy may be defi ned as including market-based

production of goods and services that are deliberately concealed

from public authorities.

Table 14 Changes in Value Added Tax rates in the euro area countries

Standard VAT rateslevel change (p.p.) change (p.p.)2007 1984-2007 2000-2007

Belgium 21.0 2.0 2.0

Germany 19.0 5.0 3.0

Ireland 21.0 -2.0 0.0

Greece 19.0 n.a. 1.0

Spain 16.0 n.a. 0.0

France 19.6 1.0 -1.0

Italy 20.0 2.0 0.0

Luxembourg 15.0 3.0 0.0

Netherlands 19.0 0.0 1.5

Austria 20.0 0.0 0.0

Portugal 21.0 4.0 4.0

Finland 22.0 n.a. 0.0

Slovenia 20.0 1.0 n.a

Euro area 19.4 1.6 0.9

Denmark 25.0 3.0 0.0

Sweden 25.0 1.5 0.0

United Kingdom 17.5 2.5 0.0

Sources: OECD. Notes: Unweighted averages for the euro area. n.a. not applicable.

50ECB


June 2008

Box 5

RECENT REFORMS OF TAX AND BENEFIT SYSTEMS. CASE STUDY: IRELAND AND GERMANY 1

Recent tax reforms implemented in Ireland have attempted to reduce the negative effects of high

marginal tax rates on labour supply. Reforms instigated in Germany between 2003 and 2005

instead sought to reduce unemployment benefi ts, tighten work availability requirements and

increase the attractiveness of employment. As these two examples show, reducing disincentives

to work, such as high marginal tax rates, high unemployment benefi ts and weak work availability

requirements can stimulate labour supply and employment.

The effect of taxation in Ireland

The tax wedge on labour was highest in Ireland in the late 1980s, fell back during the 1990s, and is

now among the lowest in the OECD countries (Nickell, 2003) and the euro area (see Section 4.1.2

above). Chief amongst these changes was the reduction in the marginal personal taxation

rate from 65% in 1983 to 41% in 2006. A narrower measure of the tax wedge that excludes

consumption taxes declined by 18.3% between 1986 and 2003 – the largest reduction recorded

in the OECD.

Higher taxes on labour contribute to higher unemployment and lower employment by reducing

labour demand (moving employers back up their demand curve) and labour supply (by moving

individuals back up their supply curves). The marked fall in these wedges during the 1990s can

be taken to have contributed to Ireland’s exceptional rate of employment creation. Research by

Callan, Van Soest and Walsh (2003) has explored the responsiveness of Irish labour supply to

various changes in the income tax code. In general, the results across different forms of tax cuts

were quite similar in many respects. However, the response of married women to a top rate tax

cut or to band-widening was more than twice as strong as that of men, and more than twice as

large as women’s response to a standard rate tax cut or allowance increase.

More recently, Callan, Van Soest and Walsh (2007) have examined the effects of increased

independence in the taxation of married couples. In 2000, the Irish tax system moved

towards greater independence in the tax treatment of couples in what has been termed an

“individualisation” of the standard rate tax band. This involved restricting the extent to which

tax bands are transferable between spouses. Callan et al. (2007) show that the impact of the

change in the tax treatment of couples on married women’s participation in the labour market

is substantially greater than cutting tax rates or increasing tax-free allowances. They estimate

that increased individualisation of tax increased married women’s participation rates by

2-3 percentage points. This increase is somewhat larger than that found by Steiner and Wrohlich

(2006) for Germany for a similar change in the tax treatment of couples.

Recent labour market reforms in Germany

Germany implemented major labour market reforms in 2003, 2004 and 2005, especially with

the so-called Hartz Acts I to IV (for an overview of these acts, see Deutsche Bundesbank, 2006,

pp. 79-83). The reform strategy included: improving employment services and redesigning active

labour market policy measures (Deutsche Bundesbank, 2006, p. 66); activating the unemployed

1 Prepared by K. Mc-Quinn and H. Stahl.

51ECB


June 2008

4 STRUCTURAL

POLICIES AND

THEIR EFFECT ON

LABOUR SUPPLY

and reducing unemployment benefi t duration; tightening work availability requirements; and

stimulating labour demand by deregulating segments of the labour market (Eichhorst and

Zimmermann, 2007; and Jacobi and Kluve, 2007).

Prior to the reform, active labour market policy measures combined with a generous benefi t

system (unemployment benefi ts amounted to 67% of the last net labour income) created strong

disincentives towards work in Germany. After entitlement to unemployment benefi t expired (after

between 6 and 32 months, depending on previous employment duration and age), unemployment

assistance was paid for an unlimited amount of time and still amounted to 57% of previous net

labour income. For a single blue-collar worker with average income in western Germany this

would have been €833 per month in 2006 in current prices.

Hartz I and II redesigned the measures for recipients of unemployment benefi ts and unemployment

assistance and many of the existing active measures, promoted low-paid part-time employment

by waiving social security contributions and deregulated temporary employment. The ineffi cient

public employment service was completely reorganised under Hartz III. Hartz IV replaced

income-related unemployment assistance with the means-tested unemployment benefi t II in

2005.2 For a single person, unemployment benefi t II amounted to about €695 per month in 2006,

net of social contributions and taxes. The unemployment benefi t I duration entitlement was

shortened to 12 months for most of the unemployed and 18 months for the elderly (> 55 years)

at the beginning of 2006. Additionally, incentives for older persons to participate in the labour

force have been strengthened.

A thorough programme evaluation is mandatory under the Hartz Acts. Micro-evaluations of

Hartz I-III indicate that the reforms of the Federal Employment Agency had some positive

effects on employment, including the self-employed.3, 4 Training programmes in combination

with training vouchers for fi rms improved employment opportunities (Schneider et al., 2007).

However, publicly-sponsored job creation schemes and public-fi nanced work agencies for the

unemployed were found to have negative effects on employment. It is too early for a conclusive

assessment of the 2003 labour market reform act and Hartz IV. However, in a recent survey, fi rms

report an increased search intensity by the unemployed, along with reduced reservation wages

(Kettner and Rebien, 2007). Moreover, unemployment fell substantially in Germany after Hartz

IV was introduced (from 2005 onwards). The effect of even small reductions in replacement

rates on employment can be sizeable.

In terms of spending, in 2006, two-thirds of the measures have yet to be evaluated. One clear

outcome of the evaluations is that the number of measures of labour market policy, which amounted

to between 60 and 80, has to be drastically reduced. Marginal tax rates for unemployed persons

moving into employment are still very high. However, since every household whose income

falls below a certain threshold qualifi es for basic allowances, reducing the benefi t-qualifi cation

level would be diffi cult. Other measures, such as “workfare”, seem more promising.

2 Income-related unemployment assistance prior to the Hartz reforms was also means-tested, but conditions were weak and referred only

to partner income.

3 The promotion of self-employment through the provision of a benefi t to unemployed workers wishing to set up their own business (paid

for 6 months and equal to the unemployment benefi t that the recipient would otherwise have received - for further details see Jacobi

and Kluve, 2007) seems to have had long-lasting positive effects on employment. Caliendo et al. (2007) report that the probability of

not being unemployed after 28 months for men in Eastern Germany is 24% higher than for comparable men that have not received this

type of benefi t. For women in western Germany, this probability is 18% higher.

4 However the current German Government is in the process of reversing some of the reforms undertaken, inter alia by increasing the

duration of the higher unemployment benefi ts for older unemployed people

52ECB


June 2008

4.1.3 PENSION SYSTEMS

Pension benefi t entitlements and their impacts

on labour supply differ widely across euro area

countries (e.g. OECD, 2007g; Blöndal and

Scarpetta, 1999; Martins, Novo and Portugal,

2007). Overall, studies point to signifi cant

negative relationships between pension benefi t

entitlement and early retirement access on the

labour supply of older workers (see, e.g. Duval,

2003 and 2006). As Table 15 indicates, the

portion of pre-retirement incomes that is

replaced by public pensions tends to be higher

for low-income workers.81 In 2004, in Greece

and Luxembourg, net pension entitlements for

low-income workers with 50% of average

earnings were even higher than previous income

from work, with net replacement rates exceeding

100%. Several reforms have been implemented

to reduce the fi nancial incentives for retiring

early. For example, for the euro area as a whole,

the implicit tax rate on continued work

embedded in early retirement schemes, which

becomes effective when a person decides to

extend employment from 60 to 65, declined by

18.2 percentage points between 1993 and 2003

but remains high.82 The reform strategies

enacted to increase working incentives for older

workers differ across euro area countries, with

most euro area countries increasing the statutory

retirement age, reducing the level of

unemployment benefi ts for older workers or

lowering the fi nancial incentives for retiring

earlier, e.g. pension levels and increases

(see Table 15 and also Box 6). Together, these

measures have been decisive in increasing the

labour supply of workers aged 55-64, as

indicated by the rise in this group’s labour force

participation rate over the last decade, as

identifi ed in Chapter 3 (see Table 2).

Further efforts to orientate tax and benefi t

systems towards increasing working incentives

and labour supply are needed. As regards

unemployment benefi t systems, these could

include improvements in benefi t administration

as well as adjustments in income support paid to

the unemployed where it reduces incentives to

search for work. Labour taxes need to be further

reduced. This can be achieved by restraining

government expenditures (e.g. through increased

effi ciency) and possibly by a partial shift from

social contributions and taxes on income to taxes

on consumption and other taxes, taking into

account redistributive effects. However, given

the impact of indirect taxes on shadow economy

workers, an overall tax reduction would most

likely be more effective for employment than a

simple shift in the tax burden for taxes on labour

to indirect taxes. As regards pension systems,

where necessary, early retirement incentives

need to be further reduced and the link between

social security contributions and benefi ts needs

to be strengthened.

Generally, a comprehensive approach to

increasing labour supply with mutually

reinforcing measures is decisive for increasing

labour supply and employment.83 By increasing

GDP, and thereby tax revenues and social

security contributions, such an approach allows

countries with sound structural fi scal positions

to further lower taxes on labour, thereby creating

a virtuous circle of raising labour supply.

Net replacement rates are estimated on the assumption that 81

individuals enter the labour market at age 20 and work until the

standard retirement age. See OECD (2007g), p.36.

Basically, the implicit tax rate on continued work measures the 82

amount of pension foregone (net of any additional contributions

paid) when an older worker decides to postpone retirement.

Since by retiring later, workers receive a pension for a shorter

period, to be incentive neutral this would need to be compensated

by a higher yearly pension in the future. If the pension increase

for retiring later falls short of compensating for one lost year

of pension, the system is said to favour early retirement. The

implicit tax rate is the amount of pension “lost” due to later

retirement (in present value terms), relative to labour income.

The implementation of tax and benefi t policies across euro area 83

countries often takes the form of policy packages, combining

either various reforms of tax and benefi t systems or linking such

reforms to reforms of other labour market institutions. The latter

is the case, for example, within the currently discussed fl exicurity

approach. Broadly speaking, this approach is supposed to combine

elements of labour market fl exibility with elements of income

security. A ‘suffi cient’ fl exibility in employees’ contractual

arrangements is meant to allow fi rms and employees to cope with

(structural) change, while employees shall be eligible to receive

an ‘adequate’ income in periods of joblessness that may arise

from the increased labour market fl exibility. Overall, the concept

of fl exicurity is intended to take account of trade-offs between

social transfers, employment protection legislation, active labour

market policies, as well as life-long learning strategies within an

integrated approach. It is decisive that such reform packages exert

a positive impact on labour supply by increasing overall labour

market fl exibility without giving rise to budgetary distortions

through higher taxes.

53ECB


June 2008

4 STRUCTURAL

POLICIES AND

THEIR EFFECT ON

LABOUR SUPPLY

Table 15 Reforms of pension systems in euro area countries

Net replacement rates 1) Implicit tax rates 2) Policy reforms 1994-2006 3)

50% of av. income level

100% of av. income level level change (p.p)

Reduced early

retirement 4)

Lower unemployment

benefi ts 5)

Financial incentives for

later retirement 6)2004 2004 2003 1993-2003

Belgium 77.3 63.0 76.5 5.9 + + +

Germany 53.4 58.0 39.0 -14.2 + + +

Ireland 65.8 38.5 37.3 -0.6

Greece 113.6 110.1 n.a. n.a. -

Spain 82.0 84.5 33.6 22.1 + [+;-]

France 78.4 63.1 50.6 -32.0 + -' + +

Italy 81.8 77.9 20.6 -79.8 + + +

Luxembourg 107.6 96.2 75.3 -3.5 + +

Netherlands 97.0 96.8 29.5 -59.8 + + +

Austria 90.4 90.9 65.6 n.a. + +

Portugal 81.6 69.2 76.5 0.9 + + +

Slovenia n.a. n.a. n.a. n.a.

Finland 77.4 68.8 59.4 -21.4 + + +

Euro area 83.9 76.4 49.8 -18.2

Denmark 132.7 86.7 n.a. n.a. + + +

Sweden 81.4 64.0 35.7 14.3 + [+;-]

United Kingdom 66.1 41.1 26.4 6.2

United States 67.4 52.4 12.8 6.5 [+;-]

Sources: OECD (2006b) Employment Outlook, Chapter 3 and OECD (2007g) “Pensions at a glance”; N. Brandt, J.-M. Bruniaux andR. Duval (2005), “Assessing the OECD Jobs Strategy: Past Development and Reforms” OECD Economics Department Working Paper No 429.Notes: Unweighted averages for the euro area. n.a. not available. Changes in the replacement rate over time are not available.1) The net replacement rate is defi ned as the individual net pension entitlement divided by net pre-retirement earnings, taking account of personal income taxes and social security contributions paid by workers and pensioners. It is displayed at 50% and 100% of average earnings, respectively. Displayed for retirement in 2004.2) The implicit tax on continued work is defi ned as the average annual change in pension/social wealth (i.e. the present value of the future stream of pensions/social benefi ts), net of additional contributions paid, resulting from a decision to postpone retirement from age 60 to age 65. The calculations are made for a single worker with average earnings. For 2003, they refl ect the steady-state of currently legislated systems thus taking account of recent reforms and their impact on future pension streams, which in some cases will take several decades. The fi gure for Belgium refers to the period 1995-2003.3) Evaluations are based on OECD (2006b, 2007g) and NCB assessments. + (-) indicates an increase (decrease) in the respective indicator or simultaneous measures acting in both directions. Assessments for Denmark, Sweden, the United Kingdom and the United States up to 2004 only.4) Includes tightened eligibility to early retirement (e.g. through increase in retirement age for early retirement).5) Includes unemployment benefi ts for older persons seeking a job with an extended duration of unemployment benefi t receipt.6) Includes actuarial adjustments to early or late receipt of pensions or fi nancial incentives to retire later (e.g. bonuses)

Box 6

ENHANCING THE PARTICIPATION RATE OF OLDER WORKERS: THE NEED FOR A COMPREHENSIVE

STRATEGY 1

Despite the fact that in most euro area countries the statutory retirement age is currently 65

for both men and women (see Table and Social Security Administration, 2006),2 in 2006, the

participation rate of individuals aged 55 to 64 stood, as for younger individuals, at around 44%.

This low rate of participation for older workers refl ects cohort effects (for women), but is also

due to economic and institutional factors. In fact, the participation rate of men between 55 and

1 Prepared by D. Nicolitsas.

2 See Table for details on the statutory pensionable age in 2005. The only euro area countries in which the statutory pensionable age

for men was not 65 in 2005 were France (60) and Slovenia (63). The euro area countries in which the statutory pensionable age for

women was less than 65 in 2005 were Belgium (64), Greece (60 but only for those women who joined the labour force prior to 1993),

France (60), Italy (60), Austria (60) and Slovenia (61).

54ECB


June 2008

64 years of age in the euro area is lower today than two and a half decades ago, while the average

effective retirement age for both men and women is lower today than in the 1980s.3

The signifi cant discrepancies between countries in the effective retirement age (see Table)

refl ect, inter alia, cross-country differences in the level and composition of economic activity,4

and in the extent of self-employment. The main factors, however, behind these discrepancies are

differences in: the statutory retirement age, early retirement schemes, restrictions on employment

during retirement, and replacement rates.5

The decline in the average effective retirement age is mainly due to the fact that in the 1980s a

number of countries tried to tackle unemployment, especially that of the young, by introducing

early retirement schemes (e.g. Belgium, Germany, Austria) and made disability benefi ts more

generous (e.g. Netherlands). At the same time, the increase in the unemployment rate amongst

older people led them to withdraw from the labour market (known as the discouraged worker

effect).

The main reason for which individuals take advantage of early retirement schemes are the high

implicit tax rates on continued work. As can be seen from Table 13, implicit tax rates are positive

in all countries and very high in some. In other words, Blöndal and Scarpetta (1999) estimate

that in the 1990s in a number of OECD countries a 55 year-old man could expect no increase in

his pension if he continued working for another 10 years – a situation in sharp contrast with what

prevailed in the 1960s.

More recently, given other developments (e.g. population ageing and the need for fi scal

consolidation), it became clear that while measures such as early retirement schemes and more

generous unemployment and disability benefi ts were not effective in lowering unemployment,6

they could, ceteris paribus, lead to a reduction in output and mount pressure on public fi nances by

reducing the size of the aggregate labour force. For this reason, a number of countries reversed

such measures and started to give employers incentives to increase the number of older workers,

and workers incentives to stay at work longer.7

Limiting the use of early retirement schemes will have wider benefi ts than just helping public

fi nances. More specifi cally, such measures could contribute to increased training activities by

lengthening the pay-back period for investment in human capital or by encouraging older workers

to weather a temporary demand shock without withdrawing permanently from the labour market

(see, inter alia, Duval, 2003). Moreover, staying in work can help older people remain integrated

in society. Despite the benefi cial effects expected as a result of the measures that have already

been planned, and the fact that the participation rate of older workers is expected to increase due

3 See Blöndal and Scarpetta (1999) for evidence until the mid-1990s and Romans (2007) for evidence on the more recent period.

4 Individuals from certain sectors in which work is more arduous (e.g. construction, manufacturing, mining and quarrying and

transportation) generally have earlier retirement dates.

5 Differences in the decision to retire early also differ across individuals depending on, inter alia, their education level; in general, there

is a negative correlation between early retirement and education level.

6 See OECD (2006d) for an illustration of the point that in 2004, countries in which participation rates of older workers were low also

had high unemployment rates (Chart 9, p.132).

7 For example, in Germany the statutory retirement age will, between 2012 and 2029, progressively increase from 65 to 67 starting

with the age cohort born in 1947; in Austria early retirement schemes are to be abolished with effect from 2017; in France a pension

supplement is given to people over the age of 60 who, despite the fact that they could earn a full pension, decide to stay on at work;

social security contributions for older employees have decreased in Belgium; in Portugal, the pension scheme reform enacted in 2007

included a bigger fi nancial penalty for early retirement – an increase from 4.5% to 6% for every year prior to the legal retirement age –

and introduced a “sustainability factor” that will make the calculation of new pensions conditional on life expectancy at 65.

55ECB


June 2008

4 STRUCTURAL

POLICIES AND

THEIR EFFECT ON

LABOUR SUPPLY

to the improvement in the educational level of the population as a whole and due to the increased

importance of the service sector, there is still need for further action. This is prompted by the

rapid expected deterioration in the old-age dependency ratio 8 (from around 25% in the euro area

in 2006 to around 35% in 2025 and 55% in 2050).9 In addition, there is still a discrepancy of

around 8 percentage points between the euro area employment rate of individuals aged 50-64

and the target of 50% for 2010 set by the 2001 Stockholm European Council. The signifi cant

cross-country differences and past experience with policy measures suggest that for the adopted

measures to succeed, it will take a comprehensive strategy that incorporates many of the special

features needed to promote the activation of older workers, not least of which is the need for

life-long learning (see, inter alia, Employment Committee, 2007). Such comprehensive reform

should also take into account other measures which may currently dampen incentives for further

training and render older employees less employable, for example those arising from the rigidity

of age-earnings profi les.

8 The old-age dependency ratio is defi ned as the population over 64 years old as a percentage of the working age population.

9 Figures from Maddaloni et al. (2006).

Statutory pensionable age and average exit age from work by gender (2005)

Statutory pensionable age

(1)

Early pensionable age

(2)

Average exit age

(3)

Statutory pensionable age

(4)

Early pensionable age

(5)

Average exit age

(6)Men Women

Belgium 65 60 61.6 64 (65) 1) 60 59.6

Germany 65 (67) 2) 63 61.4 3) 65 (67) 2) 63 61.13 3)

Ireland 65 - 63.6 65 - 64.6

Greece 4) 65 60 62.5 60 4 55 61

Spain 65 - 62 65 - 62.8

France 60 - 58.5 60 - 59.1

Italy 65 - 60.7 60 - 58.8

Luxembourg 65 57 59.4 65 57 59.4

Netherlands 65 - 61.6 65 - 61.4

Austria 65 62.5 5) 60.3 60 5) 58.5 5) 59.4

Portugal 65 55 62.4 65 55 63.8

Slovenia 63 - n.a. 61 - n.a.

Finland 65 62 61.8 65 62 61.7

Euro area 6) 64.2 60.3 61.5 62.6 58.9 61.2

Denmark 65 60 61.2 65 60 60.7

Sweden 67 61 64.3 67 61 63

United Kingdom 65 (68) 7) 50 (55) 7) 63.4 60 (68) 7) 50 (55) 7) 61.9

United States 65 62 n.a. 65 62 n.a.

Sources: DG Employment, Active ageing country fact sheets (for columns: 1, 3, 4 and 6) and Social Security Administration, 2006 for columns 2 and 5. 1) The statutory retirement age will, from 2009, increase to 65.0 for women. 2) The statutory retirement age will gradually (between 2012 and 2029) increase to 67, starting with the age cohort born in 1947. 3) Data refers to 2004. 4) In Greece, data refer to the modal case of the main pension provider for private sector employees. Statutory retirement ages are considerably lower for public sector employees, while they are less generous for farmers and the self-employed. The statutory pensionable age is 65 years for women who joined the labour force since 1993. The median exit age for women was 58.4 in 2006. 5) Early retirement is being increased gradually (by one month every four months) until it equals the statutory age – i.e. until 2017. 6) Average for the countries for which data are presented in the Table. 7) The statutory pensionable age will increase from 65 to 68 by one year in each of the years 2024, 2034 and 2044; statutory pensionable age will be equalized for men and women in the period 2010-2020; and early retirement age for a non-state pension will increase from 50 to 55 by 2010.

56ECB


June 2008

4.2 PARENTHOOD AND PARTICIPATION 84

As discussed in Chapter 3, the labour market

participation and employment rates of females

are still low compared with those of males.

According to recent calculations, euro area

GDP would be substantially increased if

female employment rates could be raised to

those of men (see Daly, 2007). Female labour

supply, that of parents in general, and fertility

developments (and thus long-term labour

supply and potential output) are substantially

infl uenced by so called “work/family

reconciliation” policies, namely childcare

opportunities, parental leave and part-time

work opportunities. These policies differ across

euro area countries and may consequently help

to explain differences in female labour supply

and fertility across countries. In general, it is

important to create incentive-compatible policy

frameworks that support both female labour

market participation and high fertility.

This section briefl y describes how these policies

affect female and parents’ labour supply and

suggests optimal policy designs for childcare,

parental leave and part-time work opportunities

for reconciling fertility and female labour

market participation.

4.2.1 FERTILITY AND FEMALE PARTICIPATION:

WHAT IS THE RELATIONSHIP?

Box 1 in Chapter 3 highlighted the importance

of fertility rates for future labour supply. Most

countries with high participation rates also have

relatively high fertility rates (such as Finland,

France and the Netherlands – see Chart 8). At

the same time, a number of countries are plagued

with both mediocre or low participation and low

fertility (Italy, Spain and Greece), while others

such as Germany and Austria have relatively

high participation rates, but low fertility rates.85

When comparing the situation in 1992 with that

of 2006, most countries retain a similar relative

position, despite the general increase in female

participation rates described in section 3.2.1.

One exception is the Netherlands (and to some

extent Ireland), which managed to move from

mediocre to become a top performer in both

categories.

Prepared by M. Ravanel and H. Schauman.84

See also Genre et al. (2005) and OECD (2001).85

Chart 8 Female participation, employment and fertility in euro area countries, 1992 and 2006 1

x-axis: participation and employment (in percentage); y-axis: fertility rate (average number of children per woman)

BE

DE

IE

GRES

FR

LUNL

PT

FIUK

IE

FIUKFR

BELU

NL PT

GR

DEES

1.0

1.2

1.4

1.6

1.8

2.0

2.2

1.0

1.2

1.4

1.6

1.8

2.0

2.2

80

participation rate (in percent)

employment rate (in percent)

denotes the average participation rate for the euro area

30 40 50 60 70

1992

1.0

1.2

1.4

1.6

1.8

2.0

2.2

1.0

1.2

1.4

1.6

1.8

2.0

2.2

7540 45 50 55 60 65 70

IEIE

BE

DEIT

GR

ES

LU NL

PTATAT

FI

IT

FI

SISI

UKUK

FR FR

BE

LUNL

PTGR

DEES

2006

denotes the average employment rate for the euro area

denotes the average fertility rate for the euro area

Source: Eurostat.Note:1) 2005 for Spain and Italy fertility fi gures.2) Countries for which no data were available for 1992 or before have been omitted.

57ECB


June 2008

4 STRUCTURAL

POLICIES AND

THEIR EFFECT ON

LABOUR SUPPLY

These different country groups refl ect differences

in policies affecting labour market participation

and fertility across countries. Policies that aim to

simultaneously increase both fertility and labour

market participation rates must focus on two

issues: fi rst, increasing the labour supply of parents

that do not yet participate in the labour market; and

second, working females (or future parents) who

are considering the fertility decision. Here, the

opportunity cost of having a child, in terms of

labour market withdrawal and possible negative

consequences for the female career, may have a

particularly strong impact on the fertility rate of

highly skilled women. Different, non-exclusive,

systems can be implemented, including the

following three possible solutions – increasing the

availability and lowering the cost of child care,

parental leave and part-time work opportunities.

Interestingly, two of the success stories, in terms

of female employment rates, namely the

Netherlands and Finland, use very different policy

mixes. In the Netherlands, fl exible working hours

and a high percentage of part-time work in total

employment appear important for both high

participation and fertility (most children do not

attend full time childcare), while in Finland, high

full-time employment and fertility is supported by

a developed childcare and extensive parental leave

system (see Box 7). The next sections will consider

these systems and their effect on participation and

fertility decisions in greater detail.86

4.2.2 AVAILABILITY AND COST OF CHILDCARE

It is widely documented that well-designed

childcare opportunities have a positive impact on

female participation (see, for example Bassanini

and Duval, 2006) and allow parents to continue

with full- or part-time work after having a child.

However, the availability and fi nancial cost of

childcare systems are important. When childcare

provided through the market is too scarce or

too expensive, one parent, usually the mother,

often stays at home to take care of the children

until they go to school. This not only keeps the

parent away from market work during the leave,

but can also affect the parent’s employment

situation after the leave due to lack of relevant

professional experience. Table 16 provides

fi gures on the type of childcare system most

frequently used by working mothers in some

European countries. Eurostat (2007) fi nds that

when no market-based childcare system is used,

the main reason given by families across the

European Union is that it is too expensive. This

can lead low-paid employees to leave the labour

market, or to accept that a large proportion of

their wage will be re-channelled into childcare.87

Note that in Finland, the price paid for childcare

is linked to family income (see Box 7).

The European 86 ad hoc module of the 2005 LFS focused on

conciliation between family and professional life. A report

issued by Eurostat in 2007 called Reconciliation between work and family life gives very interesting insights on the availability

of those different solutions across the European Union. We will

refer to their main results, especially for childcare availability.

Increasing childcare opportunities is in line with one of 87

Barcelona targets. The European Council stated that “Member

States should remove disincentives to female labour force

participation and strive, taking into account the demand for

childcare facilities and in line with national patterns of provision,

to provide childcare by 2010 to at least 90% of children between

3 years old and the mandatory school age and at least 33% of

children under 3 years of age”.

Table 16 Principal type of childcare used by employed mothers during hours of work

(this refers to children under 15, where at least one child is under 6 years old, in percent)

Market child Family, friend,care system Mother Partner Neightbour

Belgium 66 3 6 25

Germany 51 6 23 20

Spain 37 22 11 30

France 46 13 11 30

Italy 35 8 20 37

United Kingdom 34 20 17 29

Total % in all 6 countries 40 15 15 30

Sources: Micheaux, S and O. Monso (2007), European LFS 2005 and ad hoc module 2005. Data are only available for the six countries presented. Note: “Market” child care system includes both public and private provision. The column "mother" includes the case where no alternative child care is available.

58ECB


June 2008

It is important to recall that there are a number

of possible means available to lower the fees

charged for market childcare systems. These

include: increasing publicly provided and

fi nanced child care opportunities; subsidising

private provision directly; or introducing a

voucher system. Given the high costs of

childcare systems together with the need for a

balanced government budget, the effi cient use

of resources is crucial. The voucher system

supports a market-based childcare system with

the benefi ts of free competition and choice, but

has generally not been used in the euro area

countries so far. Private efforts and investments

associated with raising and educating children

have positive “external” effects for society (e.g.

in the form of future tax payments). These can

justify certain subsidies for families raising

children. However, direct child-related transfer

payments to parents may decrease incentives to

participate in the labour market. The Austrian

extension of the “Kindergeld” (child allowance)

scheme in 2002 is an example of such an effect 88

and may help to explain a negative relationship

between fertility and labour market participation

at the individual level.

For more information, see OECD (2007a). In 2002, the Austrian 88

government extended the period of entitlement for the so-called

“Kindergeld” scheme from 18 to 30 months. The maximum

duration can be extended by six months if the partner also

decides to take part in the program. Men, however, rarely

take the opportunity. The generosity of the scheme (it pays

approximately €430 per month) and the rather low threshold

for any work-related income induce most women to stay at

home for quite a long time. Empirical research has shown

that the Kindergeld scheme signifi cantly decreases women’s

participation rates. It has also been criticised on the grounds that

it is likely to lower women’s re-entry prospects in the labour

market. Critics also argue that the subsidy should be in the form

of childcare vouchers rather than a money transfer. Recently,

the Austrian government has enacted a reform to make the

Kindergeld more fl exible. Starting from 2008, parents may

chose between different lengths of payment and thus trade off

benefi t duration and monthly transfer payments.

Table 17 Length of parental leave and level of benefits

Country Maternity leave duration, weeks

Parental leave duration, weeks

% of salary during maternity leave

Paternal leave duration, days

Austria 16 100

Belgium 15 82 1)

Finland 17.5 26.5 70 2) 18

France 16 100 11

Germany 3) 14 60 (52) 100

Greece 17 100 7) 2

Ireland 18 80

Italy 21 80

Luxembourg 16 26 8) 100 2

The Netherlands 16 100

Portugal 4) 20 5 100

Slovenia 5) 15 37 100 15

Spain 6) 16 (10) 100 2

Sources: Social security programmes throughout the world, 2006 and information from the national central banks. The periods of leave listed here refer to leave granted in connection with childbirth, which usually includes a period of leave prior to and after the birth of a child. Maternity leave refers to the weeks of leave from work which can only be taken by mothers. Parental leave refers to leave that can be taken by either the mother or the father of a child. Paternal leave is optional leave that can be taken by the father in order to allow both parents to be on leave at the same time - the mother on maternity or parental leave and the father on paternal leave. In general, leave systems are complex. This table attempts to summarise the key points of the systems in place in each country. 1) For 30 days, 82% of salary; thereafter, 75%. 2) Payment up to a ceiling of €28,402 plus 40% of daily earnings for annual earnings between €28,404 and €43,698, and 25% of daily earnings for annual earnings of €43,699 or more. The minimum daily benefi t is €15.20. 3) At the beginning of 2007, the so called “Elterngeld” – a subsidy to parents who leave their job to take care of a child - was introduced. The maximum duration of 14 months requires that the leave period is shared between parents, with a minimum leave period of at least 2 months for each parent. Part-time work up to 30 hours a week is possible. 4) If the maximum leave period of 25 weeks of maternity and parental leave is opted for, the percentage of salary paid for the whole period falls to 80%. 5) The parental leave in Slovenia is known as a childcare benefi t. 6) Of the 16 weeks of parental leave, the mother must use 6; the remaining 10 can be used by either the mother or father. In 2007, paternal leave increased to 15 days. 7) 50% of salary is covered by the maternity benefi t, but mothers receive 100% of their salary while on maternity leave. 8) 13 weeks can be taken by the mother and 13 weeks by the father of a child up until the child’s 8th birthday. This is generally unpaid leave, although in certain collective agreement arrangements may be partly paid.

59ECB


June 2008

4 STRUCTURAL

POLICIES AND

THEIR EFFECT ON

LABOUR SUPPLY

4.2.3 MATERNAL AND PARENTAL LEAVE

OPPORTUNITIES

Parental leave exists in every country of the

euro area; however, benefi ts and their

duration differ from one country to another

(see Table 17).89 Maternal and parental leave

schemes generally tend to positively affect the

labour supply of females and parents

(Jaumotte, 2003). Having the right to suspend a

work activity, with the option to come back to

work afterwards without suffering a wage loss

during the leave period, can have a positive

infl uence on the decision to have a child.

Furthermore, parental leave may help

redistribute the time needed for childcare more

evenly between parents. However, Genre et al

(2005) show that parental leave has a positive

effect on employment as long as the leave period

is for less than 10 months. In addition, parental

leave schemes have been found to incur a cost

in terms of (mainly female) career progression 90

and to the extent that they are fi nanced by fi rms,

may impose a cost on employers.91

4.2.4 PART-TIME WORK OPPORTUNITIES

Part-time work is a further policy option to

conciliate family and professional lives, under

certain conditions. Indeed, some evidence in the

literature points to a positive impact of part-time

work on employment 92 and Section 3.3.2 has

highlighted the increase in part-time work since

the early 1990s, which coincided with a strong

increase in female labour supply.

At fi rst sight, working part-time seems to offer

a popular opportunity to arrange family and

work, since women choose to work part-time

far more frequently (as can be seen in Chart 9).

However, although part-time work opportunities

increase labour market participation, this does

not necessarily result in higher employment

when measured in average weekly (or annual)

hours of work (see Section 3.3.2). As shown

in Chart 9, rates of involuntary part-time work

are relatively high for females.93 For parents,

high costs or unavailability of childcare may

force workers to work part time rather than

full time. Furthermore, part-time work is

estimated to have a negative impact on career

progression.94 Balancing the rights of part-time

and full-time workers, allowing fl exibility for

both fi rms and employees through the equal

access of all working groups to part-time work

and discouraging the negative stigma that can

sometimes be associated with working part

time in some cases should be benefi cial for all

parties. Furthermore, favouring fl exible working

arrangements and teleworking could provide an

alternative to part-time work for reconciling

family and professional life.95

The information presented in Table 17 refers to parental leave 89

related to the birth of a new child, not leave for childcare more

generally.

See Mincer & Polachek (1974), Gronau & Weiss (1981) and 90

Stoiber (1990).

Possibly introducing a bias by fi rms against the hiring of 91

women.

Genre et al. (2005) document that part-time work positively 92

and signifi cantly affects the labour supply of the young, and

Pissarides et al. (2003), inter alia, emphasise the positive effects

of part-time work on female labour supply.

Generally, part-time workers have been shown to suffer 93

detrimental effects of their shorter working week on career

progression (see for instance International Labour Review

(1997), or Tam (1997)).

See for instance International Labour Review (1997) and Tam 94

(1997).

For example, in the Netherlands, fl exible working arrangements 95

and a low rate of (involuntary) part-time work coincide.

Chart 9 Part-time workers in euro area countries in 2006 as a percentage of total employment, including involuntary part-time.

(in percent)

0

10

20

30

40

50

60

70

80

0

10

20

30

40

50

60

70

80

part-time

including involuntary

mf mf mf mf mf mf mf mf mf mf mf mf mf mf mf

m males

f females

1 Euro area

2 Belgium

3 Germany

4 Ireland

5 Greece

6 Spain

7 France

8 Italy

9 Luxembourg

10 Netherlands

11 Austria

12 Portugal

13 Slovenia

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

14 Finland

15 United Kingdom

Source: Eurostat. Notes: Data on involuntary part-time work are not available for males in LU and the euro area. In these cases, only data on voluntary part-time employment as a percentage of total employment is shown (by the dotted histograms).

60ECB


June 2008

Box 7

REFORM OF CHILDCARE AND FLEXIBLE WORKING ARRANGEMENTS. CASE STUDY: FRANCE AND FINLAND.1

This box examines one important determinant of female labour market participation, namely

the childcare and fl exible working arrangement schemes in two euro area countries, Finland and

France. Finland experiences the highest female participation rate in the euro area and a high

fertility rate. France also has very high fertility rates but has female participation rates below

65%. This box gives an overview of the Finnish system and presents the most recent reform, the

2004 PAJE-reform, in France.

In Finland, a record 67.3% of women worked outside the home in 2006, compared with 71.4%

of men. Most mothers of young children work full time and only 19.2% of working women work

part-time, compared with the euro area average of around 35%. Some of the explanations for

the high female employment rate can be found in a relatively successful mix of fl exible working

arrangements for the parents of young children and effi cient childcare services.

The childcare system in Finland started in 1973 with a law on childcare encouraging

municipalities to provide childcare for all children up to the age of 6. In contrast, from 1990

onwards, municipalities were obliged to arrange childcare for all children up to the age of 3, and

since 1996 the same is true for all children below school age (the year the child turns 7). Roughly

50% of children below school age use the municipality childcare service. Of these, 77% are in

full-time day care. Only 3.5% of all children in childcare attend private childcare. There are also

part-time childcare services and around-the-clock childcare for children whose parents work in

shifts. The cost of childcare borne by parents is linked to family income - the highest fee is €200

per month, falling with each additional child. Low-income families do not pay childcare fees.

The Finnish family leave system builds on the principle that both parents should have equal

opportunities to take part in the caring of a child. For each newborn child, family leave is a

maximum of 263 days. This is split into maternal leave (105 days) and parental leave (158 days).

In addition, there is a short paternal leave (18 days). The fi nancial support paid to parents during

the leave period is tied to regular income and is at least €15.20 per day. Paternal leave allows

the father to stay at home, together with the mother, when the child is born. The popularity of

paternal leave has grown steadily, with 69% of all fathers using their right to paternal leave in

2005. After the parental leave period, one parent is allowed to get an unpaid leave from his or

her job for nursing his or her child until the child turns 3. The monetary support for one child is

roughly 300 euros per month but grows with the number of children.

In France, the fertility rate is high (at 1.93 children per woman), but female labour market

participation remains signifi cantly below that of males. Before 2004, allowances aimed to

increase fertility by lowering the “cost” of having a child. In 2004, the allowances designed to

raise fertility were reformed in order to also raise female participation.

The French allowance scheme, PAJE, consists of a birth grant, a monthly allowance paid until

a child’s fi rst birthday and two allowances for the “free choice” to reduce working time (paid to

parents who stop or reduce their paid employment to take care of their child) 2 and of third-party

1 Prepared by M. Ravanel and H. Schauman.

2 A very similar system was implemented in Germany in 2007, where parents can, under certain conditions, receive an allowance.

61ECB


June 2008

4 STRUCTURAL

POLICIES AND

THEIR EFFECT ON

LABOUR SUPPLY

4.3 IMMIGRATION POLICIES 96

As argued in Chapter 3, immigration can help to

fi ll skill shortages, contributing to the reduction

in wage pressures stemming from an increased

demand for particular skills. This suggests that

an effective immigration policy should take

into account the skills demanded in the labour

market. As discussed in Box 2, migrants can

bring skills which natives lack and cannot

acquire to a suffi ciently large scale in the short

run. Accordingly, a number of countries utilise

migration policies that select the type and amount

of immigrants let into a country over a specifi ed

time period, and design mechanisms that attempt

to facilitate the rapid integration of immigrants

with a more permanent status. Moreover, at least

in the short run, the entry of a sizeable number of

foreign workers may help to alleviate the reduction

in the size of the labour force in the near future

(see Box 1), provided that markets are suffi ciently

fl exible and policies suffi ciently effi cient to

ensure their smooth integration into the labour

market. However, there is general agreement that

immigration alone clearly cannot solve the longer-

term problems of Europe’s ageing population on

the pension and health care schemes.97

The contribution of immigration to labour

supply is regulated through immigration policies.

Traditionally, most EU governments have

selected the type of immigrants and length of

residence that suit their country’s needs. Usually

there is an annual limit (quota) for non-EU and

some EU immigrants who fulfi l a selection

criterion. As is evident from Chart 7, labour

demand has encouraged an increase in the quotas

since the beginning of the 1980s in euro area

countries.98 Furthermore, immigrants from the

EU15 99 have the right to freely move across the

euro area 100 and the EU enlargement in 2004

contributed to the increase in immigration within

the EU. Countries have tended to recognise rights

regarding asylum seekers, the marriage with a

Prepared by A. Lacuesta.96

European Commission (2002a). This conclusion is also evident 97

in the Council Opinions on the Stability Program Updates of

many European countries.

As it is further explained in OECD (2007e) (also see Box 2), 98

during the 1980s migrants generally concentrated in traditional

immigration countries (mainly Germany and Austria), while in

the 1990s, the number of immigrants increased the most in new

immigration countries (Spain, Ireland and Italy).

That is, Belgium, Germany, Ireland, Greece, Spain, France, 99

Italy, Luxembourg, the Netherlands, Austria, Portugal, Finland,

Denmark, Sweden and the United Kingdom.

The free movement of persons is one of the fundamental rights 100

guaranteed to EU citizens, and includes the right to work and

live in another EU Member State. Temporary restrictions on

the mobility of migrants to and from a new EU Member State

are possible for up to seven years following EU enlargement.

Such restrictions have been in place for some non-EU-15 EU

countries since the 2004 and 2007 EU enlargements. Moreover,

regulations such as minimum wages effectively undermine free

movement.

childcare (which pays part of a childminder’s wage). Since 2004, the aim has been to develop

participation through part-time work. Thus, the two allowances paid to part-time working parents

have been increased. This has had a positive effect on participation, but 60% of parents still choose

not to work at all. 98% of them are women, and 84% of these are blue collar workers. 37% of parents

who chose to stay at home say that another choice would have been too expensive (CNAF, 2006).

Thus, participation of low-paid mothers has not dramatically risen since the PAJE reform. Crèches

(in other countries also known as nurseries) remain the cheapest system. However, availability in

crèches is low and the opening hours do not cover a complete working day. Because availability is

so scarce, crèches are the principal childcare system for only 8% of children under 3 years of age.

In sum, the availability of affordable (e.g. subsidised) formal child care services and municipal

day care services, including crèches, in general has a positive impact on female participation

(Jaumotte, 2003; Gustafsson and Stafford, 1992; and Viitanen, 2005). Allowance packs such as

PAJE can, on the other hand, have negative effects on participation, especially of low-skilled

women. Increased availability of municipal day care services, or possibly a voucher system to

pay for private day care services, would allow for better work/family reconciliation and support

future female labour market participation.

62ECB


June 2008

foreign partner, adoption of foreign children and

reunifi cation of a foreigner’s family. This last

mechanism is an endogenous source of growth for

migration. In traditional European immigration

countries, immigrants from family reunifi cation

account for more than half of the total annual

infl ow of immigrants (OECD, 2006b).101 In

addition, income differences between origin

and destination countries combined with the

fall in transportation costs have encouraged

illegal immigration. Increased cooperation

among Member states and with countries of

origin is needed to decrease illegal migration.

Furthermore, fi rms need to be discouraged from

illegal hiring of migrants (since a large fraction

of undocumented migrants are immigrants that

have over-stayed their legal residence rather

than new entrants; OECD, 2007e). In sum, the

abovementioned factors emphasise that active

immigration policies have some limitations

in immigrant selection; however, many euro

area countries have selected their immigrants

according to employment-based migration

policies described below.

Employment-based migration policies aim to

attract particular skills in short supply on the

national labour market. Usually employers are

responsible for hiring directly, strengthening the

link between immigration and the needs of the

national labour market through ensuring a job for

the immigrant upon arrival. Governments typically

have to approve a migrant’s entry following

consideration of the country’s labour market

situation.102 Since the system is intended to solve a

limited shortage, the visa awarded is usually

temporary, although in some cases it is possible to

apply for permanent residency after several years

of residence in the destination country. The system

works fairly well for skilled migrants, since

entrepreneurs do not generally experience

problems in fi nding good candidates from abroad.

However, fi rms are more likely to fi nd unskilled

workers from the pool of already residing migrants

- encouraging illegal migration and over-stays of

unskilled workers (OECD, 2007e). In this regard,

it is important that public agencies reduce the costs

of hiring workers from abroad. Another

particularity of the system is the strong link

between the immigrant and the job held, since the

visa is cancelled as soon as a particular job is over.

The migrant cannot move from fi rm to fi rm to fi nd

a better match, and this potentially limits the

productive output of the migrant and creates

incentives for fi rms to offer lower wages to

immigrants relative to natives (due to their reduced

mobility and thus bargaining power). In order to

avoid this situation, some public supervision of

potential discrimination against foreign workers

may be required.

One alternative system, not currently

implemented in euro area countries, is the point

system. Such a system is in place in Canada,

New Zealand, Australia and Switzerland and is

being discussed in the UK. This system seeks

the admission of mainly long-term immigrants.

Immigrants are screened on the basis of a

set of characteristics that are considered to

make them more likely to assimilate in the

destination country.103 The system has been

found to be successful in attracting higher

percentages of highly skilled migrants relative

to other countries without such a system.104

One important drawback, however, is that the

national government needs to verify ex-ante

the characteristics of the immigrant. This is

especially diffi cult with regard to characteristics

such as education, where there is no good

system for comparing and evaluating the relative

quality of foreign degrees.105 Moreover, such a

system may reduce the fl exibility with which

Notice that family reunifi cation affects the quantity but also the 101

quality of immigration since the skills that are sought with a

particular selection policy do not need to apply to the individuals

who are reunifi ed.

In most cases, governments set additional quotas by region 102

and occupation once they have identifi ed particular shortages

(for example, France and Spain have a “shortage occupation

list” as opposed to a general cap as in Italy). Usually, there are

diffi culties with matching ex-ante and ex-post needs (see OECD,

2007e).

The system awards points to an individual’s characteristics 103

(e.g. education level), with a minimum number of points

required to allow an immigrant to enter the country.

Baker and Benjamín (1994) and Borjas (2001).104

In order to overcome this problem, some countries such as the 105

United Kingdom have incorporated the salary at origin of a person,

depending on the age and educational degree, in their valuation.

This information is expected to capture the different productivity

of the person, although unfortunately it also links the allocation of

visas to the origin country labour market behaviour.

63ECB


June 2008

4 STRUCTURAL

POLICIES AND

THEIR EFFECT ON

LABOUR SUPPLY

fi rms are able to hire immigrants in the face of

labour shortages.

4.3.1 POLICIES AFFECTING IMMIGRANTS ONCE IN

A COUNTRY

INTEGRATING LEGAL IMMIGRANTS

A number of studies have found that immigrants’

skills tend to be underutilised upon arrival

(Chiswick and Miller, 2005; see also Box 2),

and during the initial years after arrival, migrants

may face higher unemployment than natives.

This poor initial performance could be due to

language problems, a limited transferability of

their human capital acquired abroad, lack of

familiarity with a host country’s institutions,

or discrimination. Moreover, depending on

the degree of integration, this situation could

perpetuate over time, affecting not only the

migrants themselves but also their descendants.

Countries have encouraged several initiatives to

help better integrate resident immigrants and to

encourage the education of their children.

COURSES DEDICATED TO IMPROVING LANGUAGE

SKILLS AND PRESENTING COUNTRY’S

ADMINISTRATIVE SYSTEM

Much research stresses the importance of

language in the assimilation of immigrants

(Chiswick and Miller, 1995; González, 2005).

Thus, many countries, e.g. Germany and

Austria, encourage immigrant participation in

language programs. In some cases, however,

there are questions regarding the effectiveness

of such courses.106 In a broader initiative,

France created an offi ce for the reception of

foreigners and migration (ANAEM) in 2005 to

facilitate the reception and integration of

foreigners. However, it is still too soon to

evaluate it. Some countries have linked

complete assimilation with the right to reside

permanently and reunify some family members.

For example, since March 2006, in the

Netherlands anyone who wants a long-term

permit must pass a test including a language

exam and questions about Dutch society. This

measure has signifi cantly increased the efforts

required for individuals to be granted

permanent residence in the Netherlands, and

according to the OECD (2007e) it is a factor

underlying a decrease in the willingness of

immigrants to enter the country.

FACILITATING THE EDUCATION AND INTEGRATION

OF IMMIGRANTS’ CHILDREN

Education and language skills of the second

or even third generation of immigrants play a

role in the future success of these individuals

in the labour market. This is especially the

case for children who enter a country at a later

stage of their childhood, since the completion

of an educational level becomes more diffi cult

with age of entry (see National Research

Council, 1995; and Kate, Bachnan and

Morrison, 2001). Moreover, poor economic

conditions and any lack of assimilation of their

parents can also negatively affect children’s

educational performance. Therefore, policies

to monitor an immigrant’s progress at school

and particularly in basic subjects such as

language and mathematics seem of crucial

importance. In order to encourage school

participation, governments have tended to

promote policies such as subsidising some

costs incurred while a child is at school

(such as transportation or food). Some have also

randomised the distribution of foreign students

among different schools in order to increase the

benefi ts of being schooled with native peers

from other socio-economic conditions.

REGULARISATION OF ILLEGAL MIGRANTS

This procedure is usually used as an exception,

but due to large numbers of illegal immigrants

in recent years, it has been recently implemented

in many European countries including Spain,

Portugal, Belgium, Greece, the Netherlands and

France, and has been debated in Germany. The

procedure usually targets certain categories of

foreigners.107 As for the positive effects of such

To improve language skills of permanent immigrants, 106

compulsory courses to learn basic German were introduced in

Austria in 2006. However, the effectiveness of such measures

are subject to heavy debate.

But its effectiveness depends crucially on the amount of red tape 107

involved and the conditions set out for regularisation, since the

compulsory presentation of many documents can affect the number

of participants. For example, the OECD (2007f) has suggested that

the relatively low participation in Greece’s regularization could

have been due to the high amount of paper work required and a

requirement that a certain number of days had been worked.

64ECB


June 2008

programs, literature documents an increase in

productivity following a change in the legal

status (Kossoudji and Cobb Clark, 2002; and

Rivera Batiz, 1999).108 A disadvantage of

regularisation programs is that they do not stop

illegal migration and might even encourage it, if

migrants believe that a government is likely to

repeat the process again at a later date.

Summing up, the economic benefi ts that a host

country can derive from immigration depend

on both the number and the characteristics

of migrants that are allowed to enter the

country, and the incentives created by national

institutional frameworks for these immigrants

to fi nd work and to integrate into society rather

than rely on social security or unemployment

benefi ts. Although immigration policy does not

ensure a perfect control over immigration fl ows,

the design of national policies and institutions

are crucial to facilitating the integration of

immigrants and their children into the labour

market and society.

4.4 EDUCATION SYSTEMS AND INCENTIVES 109

The design of the education system, incentives

to invest in education and the amount and

effi ciency of national resources devoted to

education play a key role in encouraging young

people and workers to invest in education

and training, and therefore in determining the

quality of the labour force. Tax and benefi t

systems affect incentives to build up human

capital, and thus the quantity and quality of

education demanded by society.110 Despite

This research fi nds an increase in migrants’ labour opportunities 108

after legalisation, allowing individuals to fi nd a better match for their

characteristics. Moreover, legalisation tends to lead to an increase in

the investment of immigrants in country-specifi c human capital.

Prepared by N. Leiner-Killinger and M. Ward-Warmedinger.109

For example, through subsidies paid for participation in education, 110

tuition fees or taxing persons with high incomes, all of which

differ widely across euro area countries (see Table 18). Generally,

the impact on overall human capital accumulation is diffi cult to

quantify, as it depends crucially on the effi ciency of the resources

spent. A highly progressive tax system associated with high tax

wedges for low-, middle- and high-income groups (here displayed

for both 67% and 167% of the average wage) is usually assumed to

reduce incentives to increase earnings and thus also to improve skills

(via time and fi nancial resources) in human capital accumulation.

Table 18 Selected indicators of tax and benefit systems affecting education incentives

Public subsidies for education in tertiary education 1)

(in % of public expenditure on tertiary education) 2004

Average tuition fees in public institutions (in USD, using PPPs) academic year

2004-2005

Tax wedgeSingle earner 4) 67% of average worker wage

Tax wedge 2)

Single earner 167% of average wage

level 2006

change (p.p) 2001-2006

level 2006

change (p.p) 2001-2006

Belgium 15.7 853 (Fl.) 658 (Fr.) 3) 49.1 -1.6 60.7 -1.6

Germany 17.9 n.a. 47.4 -0.3 53.8 -1.2

Ireland 14.8 no tuition fees 16.3 -1.1 34.2 -1.6

Greece 5.2 n.a. 35.4 0.3 47.7 3.4

Spain 7.8 668 to 935 35.9 0.6 42.6 0.8

France 7.9 160 to 490 44.5 -3.1 53.2 1.2

Italy 16.7 1,017 41.5 -1.2 49.8 0.0

Luxembourg n.a. n.a. 30.6 -0.6 43.5 -1.4

Netherlands 27.0 n.a. 40.6 1.7 46.0 4.5

Austria 18.1 837 43.5 0.6 50.7 0.1

Portugal 5.4 868 31.7 -0.5 41.7 0.3

Finland 16.7 no tuition fees n.a. n.a. 49.9 -2.2

Slovenia n.a. n.a. 38.9 -2.5 n.a. n.a.

Euro area 13.9 38.0 -0.6 47.8 0.2

Denmark 30.3 no tuition fees 39.3 -1.2 49.5 -1.5

Sweden 28.2 no tuition fees 46.0 -1.8 54.6 -0.5

United Kingdom 23.9 n.a. 30.4 2.3 37.6 2.3

Sources: OECD (2006a), “Education at a glance” and OECD (2007h), “Taxing wages 2005-2006”. Notes: Unweighted averages for the euro area. n.a. not available. 1) Public subsidies for education to households and other private entities as a percentage of total public expenditure on tertiary education including student loans, scholarships and other grants to households. 2) See footnote 1 in Table 13. 3) Academic year 2003-04. 4) See also Table 13.

65ECB


June 2008

4 STRUCTURAL

POLICIES AND

THEIR EFFECT ON

LABOUR SUPPLY

marginal reductions in tax wedges in some euro

area countries (see Table 18), for the euro area

as a whole, incentives to increase the quality

of labour supply as measured by this indicator

have generally remained unchanged.

As discussed in Chapter 3, investment in the

quality of tertiary education is also important

for innovation and research, to the benefi t of

countries’ growth potential (see also Box 3).

Table 19 presents information on euro area

education expenditure. It shows that all euro

area countries increased their education

expenditure over the last decade, in line with the

increase in educational level of labour supply

described in Chapter 3. While expenditure

per student in non-tertiary education rose by

41% on average in the euro area between 1995

and 2004, expenditure per student in tertiary

education rose less, by 24%, partly due to

expanding student numbers. By 2004, 1.2% of

euro area GDP was spent on tertiary education,

with the vast majority of funds coming from

the public sector, and annual expenditure per

student was highest for tertiary level students.

However, comparison with fi gures for the

United States suggests that the euro area spends

a signifi cantly smaller proportion of annual

GDP on tertiary education, predominantly due

to far fewer funds from the private sector, and

that the level of expenditure per student in the

euro area is generally lower, particularly so at

the tertiary level. Furthermore, it is very

important how effi ciently money is spent 111, and

to what extent education expenditure affects the

public purse or is funded by the private sector.

Recent work by Aghion et al (2007) suggests a

signifi cantly positive link between a university’s

Chart 13 in Annex 3 shows the lack of a clear correlation 111

between aggregate Pisa scores and annual expenditure per

student for euro area countries.

Table 19 Expenditure on education

Change in expenditure on educational insitutions for

all services per student (1995 to 2004) 3

Expenditure on educational insitutions for

tertiary education as a % of GDP in 2004

Annual expenditure per student in euros 2004 4

Primary,secondary and post-secondary Tertiary Public Private

Primary education

All secondary education

All tertiary education including

R&D activities

Primary to tertiary education

Belgium n.a n.a 1.2 0.1 5267 6152 9398 6364

Germany 105 107 1.0 0.1 3927 6015 9726 6192

Ireland 181 126 1.0 0.1 4303 5643 8104 5328

Greece 2 192 151 1.1 n.a 3647 4137 4439 4075

Spain 136 167 0.9 0.3 3940 5318 7443 5237

France n.a n.a 1.2 0.2 4033 6934 8467 6254

Italy 1), 2) 105 130 0.7 0.3 5865 6225 6129 6129

Luxembourg n.a n.a n.a n.a 10681 14187 n.a n.a

Netherlands 136 101 1.0 0.3 4938 5985 10989 6348

Austria n.a 122 0.8 0.8 6087 6476 11140 7780

Portugal 1), 2) 154 98 0.9 0.1 3715 4895 6144 4610

Finland 122 110 1.7 0.1 4429 5906 9925 6189

Euro area 141 124 1.0 0.2 5069 6489 8355 5864

Denmark 121 123 1.8 0.1 6413 7023 12083 7751

Sweden 117 99 1.6 0.2 5928 6380 12871 7210

United

Kingdom 120 93 0.8 0.3 4715 5627 9114 5770

United States 130 132 1.0 1.9 6988 7887 17838 9597

Source: OECD (2007d), “Education at a Glance”. Note: euro area average is unweighted.1) Data on annual expenditure per student refer to public expenditure only. 2) Data on change in expenditure per student refer to public expenditure/institutions only. 3) Index of change between 1995 and 2004 (Expenditure is expressed in 2004 constant prices, defl ated by GDP defl ator; values represent an index with 1995=100). 4) Converted using PPPs for GDP, based on full-time equivalents, converted from US dollars to euro at January 2004 rates.

66ECB


June 2008

level of private funding and its autonomy in

spending its budget on research performance.

Some studies (for instance, Hanushek and

Luque, 2003) fi nd little or no evidence of a

positive link between more resources being

allocated to the education system and test

performance. The OECD points to the existence

of relevant ineffi ciencies in public spending on

(secondary) education. In other words,

governments could either provide the same level

of education outputs with less public spending

or, for the existing level of public spending,

increase the education sector’s performance and

effi ciency. It seems that much can already be

achieved if existing public funds are used more

effi ciently and if incentives for private funding

are enhanced.

Chapter 3 has shown that labour market

participation rates for young people are

generally low while they invest in their

education, and that the participation rate of

higher educated individuals is generally high.

This suggests that a lower participation rate of

the younger group does not present a problem,

provided that this is due to time invested in

quality education that both maximises their

chances of direct entry into the labour market

and facilitates a successful future career.112

Alongside the amount and effi ciency of

expenditure on educational systems in Europe,

an improved organisation of the educational

process is of key importance to ensure the best

quality outcome and facilitate the transition

from education to working life. Organisational

aspects include staff-to-student ratios, total

hours of tuition, the organisation of these hours

across semester and the mixture of study and

work experience. Most important in this context

are governance issues and the incentives created

by educational systems for directors, teachers

and professors to invest into the human capital

and skills of students. Wößmann (2007) fi nds

that student performance is better in countries

where private schools increase competition; in

schools that have the freedom to make

autonomous process and personnel decisions,

combined with a system of external exams that

hold schools accountable for their performance;

and in schools where teachers have both the

freedom and incentives to select the best

teaching methods. Furthermore, individual

incentives to participate in higher education,

vocational training and lifelong learning are

infl uenced by the costs of education relative to

returns (see also Box 9), which include the

expected length of time spent in education,

tuition fees and access to fi nance. Also, on the

job training (Box 8) is expected to contribute to

improvements in productivity and thus could

potentially explain differences in developments

across countries. Some reforms to support

vocational training, life long learning and

education have been undertaken in recent

years.113 Although the availability of data on

these characteristics is somewhat limited,

Table 20 shows considerable cross-country

variation in both staff-to-student ratios and

typical graduation dates across the euro area

countries. On average however, tertiary

education classes tend to be somewhat larger

and typical graduation ages higher than

comparable fi gures for the United States.114

Other institutions that matter for the effective

quality of labour supply include wage-setting

institutions (to ensure appropriate private

return) and the fl exibility of labour contracts

(as determinants of the appropriate allocation of

human capital).

The Bologna process (or Bologna accords)

aims to create the European higher education

area by making academic degree standards and

quality assurance standards more comparable

and compatible throughout Europe. Through

Information available in the EU-LFS on student status 112

suggests that the overall (for 15-64 year olds) participation and

employment rates would be 6.3 percentage points higher at

77.0% and 71.0% respectively (instead of the observed 70.7%

and 64.7%) if all inactive students were to work. The effect of

the inactive student population is larger for 15 to 24 year olds

and would increase the participation and employment rates

of this group by 31.4 percentage points to 76.0% and 68.6%

respectively (instead of the observed 44.6% and 37.2%).

For details, see the European Commission’s LABREF database.113

In comparing graduation ages, one should keep in mind that in 114

Germany, compulsory military or civil service generally takes

place after school and increases the age at which individuals

start their studies, and thus also tends to increase graduation

ages by up to one year.

67ECB


June 2008

4 STRUCTURAL

POLICIES AND

THEIR EFFECT ON

LABOUR SUPPLY

the implementation of the Bologna process,

higher education systems in European countries

should be organised in such a way that: (a) it

is easy to move from one country to another

(within the European Higher Education Area)

for the purpose of further study or employment

and (b) the attractiveness of European higher

education is increased so many people from

non-European countries also come to study

and/or work in Europe. The European Higher

Education Area aims to provide Europe with a

broad, high-quality and advanced knowledge

base; and promote greater convergence between

the United States and Europe.

In summary, educational systems and the

effi ciency of national resources devoted to

education play a key role in ensuring a smooth

transition from education to working life and

providing the labour force with relevant skills

for the future, for innovation capacity and thus

for overall labour quality within the euro area.

It is important that national education systems

are well funded and effi cient and provide

positive incentives for young people and

workers to invest in education and training,

and for directors, teachers and professors to

invest into the human capital and skills of

students. Introducing elements of competition

and external tests or exams may enhance

such incentives. It is important that such

measures also include incentives to enhance

the education of pupils with lower skills or

fewer talents.

Table 20 Ratio of students to teaching staff in educational institutions and typical graduation ages

Ratio of students to teaching staff in educationalinstitutions (2005) 1)

Typical graduation ages intertiary education

Primary Secondary Tertiary

Tertiaryeducation

programmes of3 to 5 years 2)

Advanced research

programmes 3)

Austria 14.1 10.9 15.3 22 25

Belgium 12.8 9.8 19.6 22-24 25-29

Finland 15.9 13.9 12.5 22-26 29

France 19.4 12.2 17.3 n.a. 25-26

Germany 18.8 15.1 12.2 25 28

Greece 11.1 8.3 30.2 21-22 24-28

Ireland 17.9 15.5 17.4 21-22 27

Italy 10.6 10.7 21.4 22 27-29

Luxembourg n.a. 9.0 n.a. n.a. n.a.

Netherlands 15.9 16.2 n.a. 22-23 25

Portugal 10.8 8.1 13.2 22 n.a.

Spain 14.3 10.6 10.6 20 25-27

Euro area 14.7 11.7 17.0 22.2 26.4

Denmark n.a. n.a. n.a. 22-24 30-34

Sweden 12.2 13.0 8.9 23-25 27-29

United Kingdom 20.7 14.1 18.2 21 24

United States 14.9 15.5 15.7 21 28

Source: OECD (2006a, 2007d), “Education at a glance”. Note: euro area average is unweighted. 1) By level of education, based on full-time equivalents, for Luxembourg public institutions only; the United Kingdom includes only general programmes in upper secondary education. 2) Tertiary-type A (ISCED 5A), data are available by duration of programme.3) ISCED 6.

68ECB


June 2008

Box 8

INCENTIVES FOR FIRMS AND WORKERS TO INVEST IN HUMAN CAPITAL ACCUMULATION AND LIFE-LONG

LEARNING 1

The importance of education for productivity and performance cannot be overemphasised. A

discussion of the relationship between the quality of human capital and growth has already

taken place in earlier sections of this report (see, for example, Chapter 2 and Section 3.5). In

those sections, the focus has been on formal education. But human capital improvements take

place not only before individuals start working but also while they are in work in the form, for

example, of continuous vocational training (CVT) and life-long learning (LLL). The average

annual hours spent on CVT in European enterprises (Table A), the percentage of the population

engaged in LLL activities (Table B) and the percentage of secondary level students enrolled

in apprenticeship schemes (Table C) suggest, however, that only a small percentage of the

population receives structured training outside formal education or combines studying with

vocational training. This is true even if one looks at only large fi rms or at younger individuals, as

suggested by the data presented in Tables A and B, despite the fact that the need for continuous

human capital improvement is becoming ever more pressing in view of rapid technological

progress and population ageing. That said, it should be noted that these measures in general,

and specifi cally those on CVT, do not include some more informal types of training such as, e.g.

learning-by-doing or on-the-job training, which may lie outside structured training programmes.

The limited extent to which structured training (which in a general sense also includes

life-long learning activities) takes place is often attributed to market failures; most training

is of a general nature, i.e. transferable across fi rms (at least within the same industry), capital

market imperfections prevent individuals from borrowing against human capital since returns

are uncertain (Becker, 1962), and it is diffi cult to enforce detailed contracts designed to ensure

the quality (and incentive-compatible fi nancing) of training. For example, fi rms have little

incentive to provide general training, since they are faced with the risk that their trained staff

may be poached by competitors who, having not incurred the training costs, are in a position to

offer higher wages. Despite these factors, however, a number of fi rms provide both fi rm-specifi c

and general training (e.g. German fi rms), while a number of individuals also participate in LLL

activities. The signifi cant diversity in the extent of training provided, both among fi rms and,

moreover, across countries (see Table A), is attributed to differences in industrial organisation

and to institutional features introduced to deal with market failures (see, inter alia, Acemoglu and

Pischke, 1998 and 1999; Finegold and Soskice, 1988; and Stevens, 2001) 2. For example, fi rms in

certain countries are encouraged to provide general training through tax breaks and the provision

of subsidies etc., while incentives are also provided to individuals to engage in life-long learning

through, for example, reductions in taxes, subsidised loans to trainees etc. Differences in the

extent of training also depend on the quality of the education system in each country, since

training builds on the skills that have already been provided in schools.

1 Prepared by D. Nicolitsas and H. Stahl.

2 More specifi cally, recently proposed models in the economics literature (see, inter alia, Stevens, 1996; Acemoglu and Pischke, 1998)

suggest that labour market frictions arising from mobility costs between jobs, the nature of skills, or search frictions in the skilled

labour market could explain why some fi rms do provide general training. Mobility costs arise since it takes time to fi nd an alternative

job or because employers providing the general training have an informational advantage over prospective employers regarding the

true productive capabilities of the workers they have trained. This gives these fi rms some monopsony power, which explains their

decision to invest in general training.

69ECB


June 2008

4 STRUCTURAL

POLICIES AND

THEIR EFFECT ON

LABOUR SUPPLY

It appears, however, as also suggested by the theory (see Acemoglu and Pischke, 1999) that

one of the most important factors in determining the extent and quality of training provided

is the existence of a monitoring mechanism. In the absence of a detailed contract between the

employer and the employee to ensure that the latter has received the training paid for, it seems

that the state usually fulfi ls this role by introducing a curriculum and setting standards (exams,

monitoring boards). One of the best examples of this kind is the German dual apprenticeship

system, a concise description of which is provided below.

The German dual apprenticeship system

This system (mainly for pupils aged 15 to 19 with lower and middle educational levels) combines

apprenticeships in a company with vocational education at a vocational school. It covers almost

all sectors of the economy; currently about 2/3 of apprentices are working in service sector

fi rms. The training period lasts for between two and three and a half years. For the practical

part, students receive training in a company for three to four days a week, while the theoretical

part, which is both general (language, politics, economics, religious education and sport) and

trade-specifi c, takes place at a vocational school. The quantity and quality of both the theoretical

and the practical parts are strictly regulated, and companies have to teach a broad spectrum of

tasks related to the particular apprenticeship (in other words there is a large amount of general

training involved). Students need a “vocational education contract” (Berufsausbildungsvertrag)

to begin an apprenticeship and are paid during the apprenticeship while fi nal exams are held,

for example, by the Chambers of Industry and Commerce and the Chambers of Handicrafts. In

Table A Hours in Continuous Vocational Training 1) courses per 1,000 hours worked (all enterprises) by size class, 1999 2)

Table B Percentage of all individuals participating in any kind (formal, non-formal or informal) of training activity by age 1)

(percentage, 2003)

CountrySize class (Number of employees) Age

20-49 50-249 250-499 500-999 1,000+ 25-34 35-44 45-54 55-64 Total

Belgium 5 8 9 9 13 Belgium 51 45 41 27 42

Germany 4 5 4 6 6 Germany 50 45 41 32 42

Ireland 8 8 9 24 7 Ireland 51 52 47 42 49

Greece 1 2 4 3 6 Greece 27 19 13 7 17

Spain 4 5 9 9 11 Spain 33 26 20 14 25

France 5 7 9 12 16 France 61 55 51 32 51

Italy 3 4 5 7 10 Italy 57 52 47 35 49

Luxembourg 3) n.a. 5 n.a. 13 n.a. Luxembourg 86 84 79 75 82

Netherlands 7 10 11 12 14 Netherlands 51 44 39 30 42

Austria 4 4 5 5 8 Austria 90 88 87 93 89

Portugal 1 3 5 10 8 Portugal 54 46 39 33 44

Slovenia 2 3 3 7 7 Slovenia 86 83 80 78 82

Finland 9 8 10 12 13 Finland 85 82 76 66 77

Euro area 4.4 5.5 6.9 9.9 9.9 Euro area 60.2 55.5 50.8 43.4 53.2

Denmark 11 14 10 13 15 Denmark 82 83 80 72 80

Sweden 10 8 11 16 14 Sweden 77 74 71 62 71

United Kingdom 6 7 14 17 6 United Kingdom* 44 42 39 23 38

United States n.a. n.a. n.a. n.a. n.a.

Source: Eurostat. Notes: n.a is not available. Euro area averages are unweighted. 1) Apprenticeship schemes are not included. 2) In January 2008, Eurostat data for a later date (2005) were available for only a few countries. From the available data, it appears that the number of hours spent on CVT courses was still low, yet higher than in 1999. 3) The fi gure for fi rms with between 500 and 999 employees refers to fi rms with over 250 employees.

Sources: Eurostat, LFS and ad hoc module on lifelong learning 2003 (see Kailis and Pilos, 2005). Note: 1) Figures refer to the whole population over the age of 25 independent of activity status. * Informal training not included for the UK. Informal training differs from non-formal training because it corresponds to self-learning through, for example, books, computers, learning centres or educational broadcasting.

70ECB


June 2008

2002, an estimated 62% of apprentices found a full-time job once they had passed the fi nal exam,

while 32% were unemployed.3

The dual apprenticeship system has a long tradition in Germany. Recently, however, certain

reform needs have been identifi ed. A signifi cant problem is that the supply of vocational

education contracts is persistently lower than demand. Only 25% of fi rms offer apprentice

opportunities. Many fi rms are too small to participate and fi rms face the risk of losing their

investment (including the sometimes relatively generous apprenticeship wages) if apprentices

are not retained at the end of their training period.4 ,5 A lack of quality applicants can also be an

obstacle. Apprenticeship schemes (356 of them) are considered overly specifi c and discussions

are underway to consider reducing them to about 40 broader occupations. Companies would

then have to teach only those skills that are central to a particular apprenticeship. In an effort to

modernise the dual apprenticeship system, the Vocational Training Reform Act was enacted on

1 April 2005.6 This act facilitates the accreditation of qualifi cations acquired outside the dual

apprenticeship system, for example at full-time vocational schools.

3 These employment and unemployment fi gures should be interpreted with great care since some agreements with trade unions explicitly

specify that fi rms must offer to those apprentices who pass the exam at least a temporary job. After two months of having completed

the apprenticeship scheme, male apprentices of German nationality do quit their job, probably in order to do their military or civil

service, although this cannot be identifi ed with the available data.

4 Monthly wages for apprentices range from €200 in transport to €1,000 for apprentices in hotels and restaurants.

5 According to a survey of 2,500 fi rms, total costs per year and per student are €16,435, half of which are labour costs. Subtracting the

gains from the students’ labour input, there remains a loss of €2,400 per year. However, the fi rm can save €5,800 in the opportunity

costs of hiring an external applicant if the student stays at the fi rm after having received his/her apprenticeship.

6 A more detailed presentation of the reform of the German dual apprenticeship system can be found in the site of the Federal Ministry of

Education and Research, “Reform of Vocational Education and Training” (http.//www.bmbf.de/en/1644.php).

Table C Percentage of upper secondary education students enrolled in apprenticeship schemes (combined school and work-related schemes), 2005

BE DE IE GR ES FR IT LU NL AT PT SI FI Euro area DK SE UK US

3.3 45 3.8 n.a. 2.8 11.3 n.a. 13.6 20 32.7 n.a. 3.7 10.5 14.7 47.7 n.a. n.a. n.a.

Source: OECD (2007d), “Education at a glance”, Table C.1.1. p. 277. Notes: n.a is not available. Euro area unweighted average calculated on the basis of available information.

71ECB


June 2008

5 MATCHING

LABOUR SUPPLY

AND DEMAND5 MATCHING LABOUR SUPPLY AND

DEMAND 115

A well functioning labour market requires a

good match between labour supply and labour

demand, which should in turn support high

employment rates. In the absence of a shift in

relative demand, an increase in the relative supply

of a particular group of workers, such as the high

skilled, will result in lower wages (i.e. returns) or

higher unemployment for those workers relative

to workers with less education. Generally,

sector specifi c shocks in an environment of

matching ineffi ciency and wage rigidities will

lead to general wage increases in excess of

productivity growth and upward pressure on

prices. An effi cient matching process, combined

with fl exible wages, should therefore reduce the

risk of upward pressures on wages and infl ation

resulting from shocks to relative labour demand

and support employment creation and output,

facilitating adjustment to monetary policy

actions and economic shocks.116

This chapter reviews the evidence on returns

to education and the relative unemployment of

particular groups in the labour market. It aims

to consider the extent of mismatches between

labour demand and labour supply in the euro

area, updating and extending some of the

analysis undertaken by the European Central

Bank (2002). It includes a brief overview of

how labour and product market institutions

affect the matching of labour supply with labour

demand. It concludes with a discussion of the

implications of globalisation and demographic

and technological change for future labour

supply. The main fi ndings of this chapter include

only limited evidence of increasing returns

to education, high rates of unemployment for

low-skilled workers in some countries (e.g.

9.3% overall, but 17.6% in Germany in 2007)

and non-EU nationals (14.7% in the euro area),

persistent labour market shortages of some skill

groups (e.g. those involved in teacher training

and education, engineering and manufacturing,

and health and welfare qualifi cations), and

high rates of youth unemployment in the euro

area (19.3% for 15-19 year olds in 2007).

Furthermore, unemployment rates vary

signifi cantly by geographical region, suggesting

a lack of cross-regional labour mobility and

insuffi cient regional wage differentiation. There

is evidence that euro area countries may have

been underutilising the increases in female

and non-national labour in recent years. The

challenges facing the euro area imply that labour

markets must not only be adequately prepared

with an appropriately enhanced quantity and

quality of labour supply, but must also rise to

the task of ensuring an effi cient match between

its workforce and labour demand.

5.1 RETURNS TO EDUCATION

Returns to education are a function of the

quality of labour (education and skills) and of

the matching of labour demand with labour

supply. Chapter 3 has shown that an increase

in the supply of highly educated workers has

been an important trend in euro area countries

over the last 20 years. In the face of skill-biased

technological change and a rapid increase in

the demand for skilled workers, empirical

studies have shown a signifi cant upward

shift in returns to education in the United

States, whereas studies for other (including

European) countries do not fi nd such a shift 117

(see Box 9). Chart 10 suggests that for the

United States, an increase in total hours worked

by the highly skilled has been accompanied by

an increase in the group’s hourly wage, whereas

in the euro area, the increase in total hours

worked by the highly skilled has, on average, not

been matched by a signifi cant increase in hourly

wage. 118 A possible explanation for this evolution

in hourly wages and in returns to education for

the euro area is that the higher relative supply of

skills was not matched by an increase in relative

demand for skills in Europe. Alternatively,

the quality of education in Europe may not

match that in the United States (as discussed in

Chapter 4), or Europeans may have tended to


See, e.g. Christoffel, Kuester and Linzert (2006).116

Ashenfelter et al. (1999).117

Although trends in total hours and hourly wages by skill group 118

vary across euro area countries.

72ECB


June 2008

invest more in particular subject areas that are

less rewarded by the market. Other possible

explanations are a lack of wage differentiation or

relative wage fl exibility between different skill

groups, relatively high taxes on higher incomes

or an increase in competitively priced labour

from abroad, which has led to an adjustment

via quantities instead of wages, thus provoking

an increase in the unemployment of low-skilled

relative to high-skilled workers. In a similar

fashion, an increase in the relative supply of

workers with more experience, for example as

the baby-boom cohort reaches prime working

age, and an increase in female labour supply both

have implications for returns and unemployment

rates of these labour input types, depending on

the existing labour market rigidities in those

particular subgroups. The unemployment rates

of different groups of workers are considered

further in the next section.

Chart 10 Hours worked and hourly real wages by educational attainment-based measure of skill

hours high skilled

hours low skilled

real wage high skilled

real wage low skilled

y-axis: 1990 = 100

x-axis: years

Euro area United States

50

70

90

110

130

150

170

50

70

90

110

130

150

170

1990 1992 1994 1996 1998 2000 2002 200470

80

90

100

110

120

130

140

150

70

80

90

100

110

120

130

140

150

1990 1992 1994 1996 1998 2000 20042002

Source: ECB calculations based on EU Klems data.Note: Skill data are derived from national data on educational attainment, with “high skilled” comprising those with university level education and “low skilled” comprising those with primary and/or secondary education (depending on the country). Total economy data, i.e. manufacturing plus services.

Box 9

RETURNS TO EDUCATION IN THE EURO AREA COUNTRIES 1

Markets provide incentives for individuals to invest in skill acquisition through offering higher

wages or earnings for higher levels of education and/or a specialisation in particular fi elds. A

lack of skill acquisition, either to a higher level or within particular fi elds, may refl ect net returns

to education that are too low. Policies that aim to increase the overall skill level, or the supply

of particular skills in an economy, must therefore be confi dent that the supply of skill will be

met by an adequate demand from fi rms, and thus that appropriate rates of return to investment in

education are in place.

1 Prepared by J. Turunen and M.Ward-Warmedinger.

73ECB


June 2008

5 MATCHING

LABOUR SUPPLY

AND DEMANDA driving factor in the trend of mean returns to education for many countries has been changing

returns to higher education. A review of available cross-country evidence on so-called

Mincerian returns to higher education, i.e. the premium in earnings for those with a tertiary

(university level) education relative to those with primary education (Budria and Pereira,

2005) or high school education (Acemoglu, 2003), suggests that returns to higher education

appear to have increased only in three out of twelve euro area countries considered since the

1980s. Thus, over the period 1980 to the mid 1990s, at least, there was no general increase in

the returns to education in the euro area, as could have been suggested by a potential increase

in relative demand for highly educated workers owing to skill-biased technological change.

A possible reason for this includes that demand was more or less met by the increase in the

supply of skills described in Chapter 3.

Individuals’ decisions to acquire human capital, in principle, involve a substantial forward-

looking component and a complex calculation weighing the costs of an additional year spent

in education against the expected benefi ts from the investment over an entire lifetime. For

example, if the alternative involves possibly longer periods of future unemployment, higher

(expected) unemployment benefi ts may contribute to reduced investment in education/

human capital. Furthermore, there may be high “opportunity costs” attached to study or work

vis-à-vis leisure for those young people with a high preference for leisure. Mincerian returns

do not take into account the cost of investing in education borne by the individual, and a

useful alternative measure of the returns to education is thus the internal rate of return.2

Although measurement diffi culties suggest that such estimates should be read with caution,

the OECD (2006a) estimates that the private internal rate of return for individuals obtaining a

university-level degree directly following an upper secondary and post secondary non-tertiary

level of education exceeds 8% in a sample of 11 OECD countries in 2003. Rates of return

were estimated to be as high as 16% in Finland and up to 15% in Belgium (see Table 33 in

Annex 3). Furthermore, this study fi nds lifelong learning to be worthwhile, as the estimated

rates of return to a 40-year-old resuming the next level of their higher education on a full-time

basis were above 6.5% in all countries in 2003, signifi cantly higher than the rates of return for

a younger student in some countries.3

National studies also show that the returns to investment in higher education vary by gender and

subject group. For example, evidence for Germany (Lauer and Steiner, 2000) suggests higher

rates of return to education at technical universities. Ammermüller and Weber (2005) fi nd higher

rates of return to law, business and medicine subjects, and lowest rates of return to education,

arts, agriculture and theology. Studies for Germany and Greece fi nd higher returns to education

for women.4 Furthermore, some countries show that within-group income inequality is increasing

for the most educated workers (Luxembourg, Finland, Austria and Portugal). In this context it

should be noted that well-designed education and high skill levels also provide non-pecuniary

returns to benefi ting individuals (e.g. social status, social contacts), as well as positive external

effects to society (e.g. through learning within social groups and integration effects). Although

2 Defi ned as the rate that equates the costs to individuals of attaining the next highest level of education with the present value of an

individual’s lifetime stream of additional earnings associated with the higher level of attainment.

3 One possible explanation for this fi nding is the existence of cohort effects in educational level. Chapter 3 has shown that over the last

two decades, the supply of workers with a university degree has increased signifi cantly. However, for older workers, higher returns to

higher education may refl ect the relatively short supply of degrees within this cohort. Similarly, investment in education that is well

matched to individual career plans and fi rms’ needs may yield higher returns.

4 See, e.g. Ammermüller and Weber (2005), Kanellopoulos et al. (2004), Magoula and Psacharopoulos (1999), Papapetrou (2007).

74ECB


June 2008

5.2 RELATIVE UNEMPLOYMENT RATES OF

DIFFERENT GROUPS OF WORKERS

This section presents evidence of mismatch

between labour supply and labour demand 119 in

the euro area and euro area countries, through

the presentation of unemployment rates for

different labour market groups – i.e. a measure

of excess labour supply 120 - and of the range of

group-specifi c unemployment rates.121 This

latter measure describes the difference between

the highest and lowest subgroup unemployment

rates. A greater range implies a larger relative

labour market imbalance. Information on the

persistence of such imbalances is acquired by

considering the range over time. Large ranges

which persist over time may be indicative of

serious labour market imbalances. This section

fi rst briefl y reviews institutional elements which

may contribute to labour market mismatch. It

then considers the extent of labour market

mismatch by: (i) Educational level and subject

area – here mismatch may occur when the level

and/or specialisation of workers’ qualifi cation

do not match those demanded by employers.

This may result in groups of workers

experiencing problems in fi nding or keeping a

job due to under- or overeducation or

inappropriate subject specialisation. (ii) Region

– here mismatch occurs when the workers in

one region experience excess demand for their

services and workers in another are in excess

supply, related to, for example, language or

other barriers which prevent cross-border and

other geographical labour mobility. (iii) Age,

gender and nationality – Chapter 2 has shown

that the largest changes in the composition of

labour supply are driven by these characteristics.

Differences in unemployment rates by age,

gender and nationality may therefore be

informative about both the impact of labour

market rigidities and the extent to which

increases in labour supply over the recent period

are met by demand. Such labour market

mismatches may refl ect both objective economic

reasons (such as adjustment costs, regulations,

skill level, specialisation, work experience) and

other factors (such as discrimination or

prejudices against certain groups).

5.2.1 INSTITUTIONS AFFECTING THE MATCHING

OF LABOUR SUPPLY WITH LABOUR DEMAND

A number of institutions may slow down or

inhibit the matching of labour supply with

labour demand. First, wages that are not

suffi ciently differentiated, for example, by skill

Two broad types of mismatch of the supply and demand for labour 119

can be identifi ed. First, although it may appear that labour supply

is available to feed existing labour demand, it may take time for

unemployed workers to fi nd jobs, given that workers do not have

full information about available positions and fi rms do not have

full information about available labour. Some amount of this

“frictional” labour market mismatch may be unavoidable, even in a

well-functioning labour market, and its size is determined by the

effi ciency of the process of collecting, processing and assessing

the necessary information for both employers and the unemployed,

and by labour market and product market rigidities. Labour market

mismatch can also be caused by “static” mismatch, where the supply

of and demand for labour do not match, for example as a result of

asymmetric information or time-inconsistent expectations by workers

of fi rms’ skill demand and/or a one-off increase in demand or supply

of, e.g. particular skills. Such one-off shocks may arise from business

cycle developments and/or from infl exible adjustment mechanisms,

e.g. through strict EPL or wage rigidity. The less fl exible markets are,

the more persistent mismatches tend to be.

Alternatively, one could consider measures of excess labour 120

demand, such as vacancy rates. However, no reliable data on excess

labour demand exists for all euro area countries. For example,

in Belgium and Luxembourg, vacancy data suffers from, for

instance, double counting, underreporting and/or selective sampling

problems. An alternative measure of labour market imbalances

could be provided by data on the number of fi rms restricted in

their activity due to labour shortages. This data is however not

available for all countries and for all sectors of economic activity.

From the available data, it appears that in the period 2004-2007,

this percentage has been increasing, and the percentage of fi rms

constrained in their activity is higher in services than in industry.

For an alternative measure based on variance of group specifi c 121

unemployment rates, see Lipsey (1960) and European Central

Bank (2002).

not usually included in estimates of the returns to education, such factors are also important for

assessing benefi ts to education.

Returns to education are important in giving incentives for young people to invest in education and

training. To ensure the future effi ciency of education and training programmes, it is furthermore

important that those investing in higher education are well-placed by ability and subject.

75ECB


June 2008

5 MATCHING

LABOUR SUPPLY

AND DEMANDor region may contribute to increasing the

mismatch between labour supply and labour

demand (e.g. by not providing the incentives for

capital to shift to the area/regions with high

unemployment or for workers to move to

regions with labour market shortages), thus

increasing the unemployment rates of some skill

groups and in some regions.122 If relative wage

compression is too strong, low-skilled workers

or workers living in low-productivity regions in

particular may remain unemployed. Similarly,

minimum wages which are too high may price

young and lower skilled workers out of the

labour market. This suggests that wages and

labour costs must adjust fl exibly to refl ect local

and sectoral labour market conditions – such as

regional unemployment rates, productivity

growth and workers’ skills.123

Second, non-wage costs that are too high, or fall

too heavily on one type of labour, may contribute

to wage compression and also increase the

labour market mismatch of certain groups. For

example, if taxes fall on employees in the form

of lower wages, this may have negative effects

on the labour supply of particularly low-wage

earners such as the low skilled. High labour

taxes also reduce the possible incentive and re-

allocation effects of a wage change.

Third, overly strict employment protection

legislation (EPL) has been found to increase

the costs to fi rms of adjusting their workforce,

reducing labour market turnover, and can create

a barrier to hiring. A number of studies have

found that overly strict EPL reduces particularly

employment of workers experiencing problems

entering the labour market, such as young

workers and women (Bertola et al 2002,

Jimeno and Rodriguez-Palenzuela 2002,

OECD 2004). This suggests that EPL should

be designed to distort labour turnover to the

lowest extent possible and coordinated with

other policies, such as unemployment benefi t

systems and active labour market policies124.

In addition, creating more individually tailored

and fl exible contracts would be advantageous125

(for example, giving workers the choice to

negotiate more or less employment protection

from fi rms, based on their individual situation

and preferences).

Fourth, the infl exibility of working hours per

employee may reduce the matching of labour

supply and labour demand, in particular for older

workers and women. On the demand side, fi rms

may prefer to adjust hours of work rather than

employment over the business cycle. Similarly,

some workers may prefer fl exible hours or

part-time work in order to combine family and

working life responsibilities (see also Chapter 4,

section 4.2) or to adjust working time with shifts

in leisure preferences or other activities over

their lifetime. Increased working time fl exibility,

which allows the combination of work and family

life and of phased retirement, may facilitate the

matching of labour supply and demand.

Fifth, product market regulation has been found

to restrict employment in the OECD countries

(Nicoletti and Scarpetta, 2005; Nicoletti et al.,

2001).126 Furthermore, the presence of start-up

costs (in particular administrative burdens on

the creation of new companies) may hinder the

growth of some sectors or industries, creating

bottlenecks in the process of matching the

demand and supply of workers in different

sectors of an economy (see Lopez-Garcia,

2003; and Rogerson, 2003). Policies to increase

competition in product markets, such as

reductions in administrative burdens on start-

ups and the removal of statutory barriers to

entry in certain sectors should help to support

employment creation through the creation of

Many studies highlight the insider-outsider theory of the labour 122

market (e.g. Amable et al., 2007; Bertola et al., 2002). According

to this theory, the labour supply of labour market “insiders” are

not so much affected by union density and centralisation and

employment protection. This means that institutions have a more

limited effect on the supply of prime-age males. In turn, they

predominantly affect young workers and females in particular,

since these are not so attached to the labour market (outsiders).

This may include giving fi rms greater scope to adjust wages to 123

these local conditions, for example through the greater use of

opening clauses and ensuring that wage fl oors such as minimum

wages are not set too high to negatively affect employment and/

or are differentiated by region or age.

See also footnote 82 for a discussion of policies to support 124

fl exicurity.

See, for example, Carnoy et al. (1997)125

Product market regulation can affect the labour supply through 126

different channels.

76ECB


June 2008

new fi rms, but also by constraining the wage rents

of insiders due to increasing competition.127

Finally, mismatches that result from a defi cit of

certain skills in the labour market may be eased

through higher investment in human capital

acquisition – through increased lifelong learning

and training and educations systems which put a

greater emphasis on identifying future skill needs.

5.2.2 EDUCATIONAL MISMATCH

In all euro area countries, the unemployment

rate decreases signifi cantly with the level of

educational attainment. Table 21 shows that in

2007, the euro area unemployment rate 128 was

9.3% for those with lower secondary education,

compared with 6.4% for those with upper

secondary education and 4.1% for those with

tertiary education. The unemployment rate for

those with lower secondary (tertiary) education

was therefore signifi cantly higher (lower) than

On the one hand, lower barriers to entry for new fi rms and 127

increased real wages due to lower prices may positively affect

activity. On the other hand, employees’ bargaining position may

be weakened by reducing wage rents. Bassanini and Duval (2006)

document a negative effect from a decline in regulation on the

employment of females. Fiori et al. (2007) document a signifi cant

negative effect from product market regulation (PMR) on overall

employment, implying positive effects from deregulation. These

authors also show that labour and product market regulation

interact in two respects. First, PMR has a larger negative impact

on employment when labour markets are also regulated. Second,

rigid PMR has facilitated labour market deregulation, but not vice

versa. Furthermore, wage moderation has a strong yet positive

impact on employment if product market regulation is low.

It should be noted that methodological changes due to, e.g. 128

census revisions, changes in concepts and in a change from the

use of annual to quarterly or more frequent surveys and data,

have resulted in breaks in the LFS survey over time in many

euro area countries, which may affect trends in unemployment

over time for some countries (see Annex 2). However, it should

also be noted that data on unemployment levels, presented in

this section for spring 2007, are well-harmonised.

Table 21 Unemployment rates and mismatch by the level of educational attainment in the euroarea and euro area countries 1)

Country

Unemployment rate in 2007 (%) Educational mismatch

Low Medium High

Weighted average of subgroups

Range 2007

Change in range between highest and lowest rate (p.p) 6)

1993-1995 2) 1996-2001 2002-2007

Belgium 11.7 6.2 3.4 6.5 8.3 2.5 0.8 -2.4 -0.4 2.0 0.3Germany 17.6 8.3 3.6 8.2 13.9 3.2 1.1 -0.2 0.0 5.2 0.9Ireland 6.4 3.5 2.3 3.8 4.1 -3.1 -1.0 -7.8 -1.3 0.0 0.0Greece 7.1 8.0 5.8 7.1 2.3 0.5 0.2 0.2 0.0 -0.8 -0.1Spain 8.4 6.5 4.7 6.7 3.7 -0.2 -0.1 -2.4 -0.4 0.1 0.0France 4) 11.2 6.8 4.8 7.4 6.3 n.a. n.a. -0.4 -0.1 -0.6 -0.1Italy 6.0 3.8 4.2 4.7 2.3 0.3 0.1 1.9 0.3 -1.5 -0.3Luxembourg 4.2 7) 2.4 7) 2.9 7) 3.1 1.8 7) n.a. n.a. n.a. n.a. n.a. n.a.Netherlands 5) 4.1 2.7 1.7 2.6 2.4 n.a. n.a. -3.4 -0.6 1.7 0.3Austria 3) 8.4 3.6 2.4 4.1 6.0 n.a. n.a. 1.2 0.2 1.2 0.2Portugal 8.2 6.7 6.0 7.6 2.2 n.a. n.a. -2.2 7) -0.4 7) 0.8 7) 0.1 7)

Slovenia 5) 7.1 7) 4.4 7) 2.7 4.3 4.4 7) n.a. n.a. 0.4 7) 0.1 7) -1.3 7) -0.2 7)

Finland 3) 8.9 5.9 3.4 5.4 5.5 n.a. n.a. -3.0 -0.5 -1.5 -0.3Euro area 9.3 6.4 4.1 6.6 5.2 1.7 0.6 -1.4 -0.2 0.3 0.1Denmark 4.0 2.4 2.9 3.0 1.6 -0.9 -0.3 -3.6 -0.6 -0.3 0.0Sweden 3) 6.9 4.3 3.5 4.4 3.4 n.a. n.a. -3.2 -0.5 0.2 0.0United Kingdom 6.0 3.6 2.0 3.6 4.0 -1.3 -0.4 -1.9 -0.3 0.2 0.0

Sources: EU-LFS (spring data) and ECB calculations. For all tables on mismatch, the fi gures presented are those fulfi lling Eurostat's publications and reliability limits for the LFS data8). In bold are the best three performers in terms of (i) a low level and (ii) a low range of unemployment rates. The weighted average of subgroups may not equal the total unemployment rate due to missing data. 1) 25 to 64 years old; the education levels refer to low: lower secondary education and less, medium: upper secondary education, high: tertiary education. 2) Education data start in 1992. 3) Data start in 1995. 4) Data start in 1993. 5) Data start in 1996. 6) The range refers to the difference between the highest and lowest sub-group unemployment rate. Average annual changes are presented in yellow. 7) Based on fi gures smaller than the Eurostat reliability limit. 8) The underlying fi gures for the mismatch indicator (based on the unemployment rate) are unemployment and employment. For every country and the euro area, these fi gures are compared with the respective Eurostat limits. Whenever the numbers are smaller than the publication limit, they are omitted from the published table (these cases are labelled n.a.). Whenever the numbers are larger than the publication limit but below the reliability limit, they are fl agged separately.

77ECB


June 2008

5 MATCHING

LABOUR SUPPLY

AND DEMAND

the total unemployment rate of 6.6%, indicating

the much greater unemployment problems

experienced by people without good academic

or vocational qualifi cations and the stronger

demand for employees with a higher level of

education, relative to the supply of such

workers.

The range in unemployment rates across

educational levels was 5.2 percentage points in

2007 for the euro area as a whole, having

narrowed in the generally favourable cyclical

environment of the late 1990s. In recent years,

there have been pronounced differences in

country-specifi c developments. In Germany,

overall unemployment increased rather steeply,

and the gap in unemployment rates by

educational level widened markedly to 13.9

percentage points in 2007 129. In contrast, Ireland,

which experienced a more favourable trend in

total unemployment, saw a decline in the gap

between the unemployment rates of low-

educated and highly educated persons in the

second half of the 1990s.

Unemployment rates also vary signifi cantly

across type of education in the euro area and

euro area countries (see Table 22). Workers

with humanities and services specialisations

typically experience unemployment rates

around or slightly higher than the national total

unemployment rate. On the other hand, workers

with teacher training and education;

engineering and manufacturing; and health and

welfare qualifi cations typically experience

In Germany, from 2005 onwards, persons capable of work 129

applying for basic social security had to register at the

employment offi ce as unemployed and actively seek work. This

might have contributed to the number of low-skilled unemployed

in the German LFS.

Table 22 Unemployment rates and mismatch by type of education in the euro area and euro area countries 1)

Country

Unemployment rate in 2006 (%) Mismatch (range)

General

Teachertraining

andeducation

Humani-ties

languages and art

Socialscience

business,law

Science math

comput-ing

Engineer-ing manu-facturing

Agricul-ture

Health and

welfareSer-vices

Weighted average of sub-

groups 2) 2006

Change (p.p) 5)

2003-2006 3)

Belgium 5.4 2.4 7.8 5.6 4.4 4.6 n.a. 3.4 8.2 7.0 5.8 -1.2 -0.6 Germany 7.3 4.3 7.6 7.4 6.5 10.4 9.1 5.2 9.2 9.8 6.1 -1.9 -0.6 Ireland 4) 2.9 1.6 4.3 1.8 2.9 2.7 0.4 1.5 2.5 3.4 4.0 n.a. n.a. Greece 7.8 4.2 7) 9.2 9.1 9.4 5.2 8.4 7) 9.2 8.1 7.6 5.2 7) -0.6 7) -0.2 7) Spain 6.7 5.1 7.8 6.7 7.2 7) 4.2 4.1 5.6 8.8 7.3 4.6 -0.9 -0.3 France 14.9 n.a. 9.3 7.2 6.1 5.9 4.5 3.7 9.8 8.1 n.a. n.a. n.a. Italy n.a. 4.5 6.5 5.2 6.4 3.3 4.2 2.5 6.0 5.6 n.a. n.a. n.a. Luxembourg 6.3 n.a. n.a. 3.2 n.a. 1.9 n.a. n.a. 4.6 4.0 n.a. n.a. n.a. Netherlands n.a. 2.1 7) 3.8 7) 2.9 3.7 7) 3.7 2.0 2.5 3.3 3.9 3.4 n.a. n.a. Austria 4.4 n.a. 5.0 7) 3.8 n.a. 3.2 n.a. 2.2 7) 4.4 4.1 n.a. n.a. n.a. Portugal n.a. n.a. 8.2 6.0 5.9 6.4 n.a. n.a. n.a. 7.2 n.a. n.a. n.a. Slovenia 8.0 7) n.a. 7.4 7) 5.9 n.a. 4.5 4.8 3.4 7) 4.5 5.3 n.a. n.a. n.a. Finland 8.0 n.a. 8.4 5.3 6.9 7) 5.1 5.5 2.0 6.9 6.3 n.a. n.a. n.a. Euro area 6) 7.1 4.0 7.7 6.5 6.3 7.1 5.7 4.3 8.0 7.5 4.0 -0.4 -0.1 Denmark n.a. n.a. 5.7 3.2 n.a. 2.1 n.a. 2.5 3.5 3.3 n.a. n.a. n.a.Sweden 6.0 2.8 7.7 5.2 n.a. 4.7 4.4 7) 3.0 4.8 5.2 4.8 -2.4 7) -1.2 7)

United

Kingdom n.a. 1.5 3.5 3.0 3.4 2.5 n.a. 1.8 5.1 3.9 3.6 0.9 7) 0.5 7)

Sources: EU-LFS (spring data) and ECB calculations. For all tables on mismatch, the fi gures presented are those fulfi lling Eurostat's publications and reliability limits for the LFS data. Notes: data stem from different LFS sources than in the case of the other breakdowns and are therefore not fully comparable. The detailed education breakdown is only available for 2006.1) 25 to 64 years old. 2) The weighted average of subgroups may not equal the total unemployment rate due to missing data. 3) Starting 2004 for AT, BE, IE, PT, SE, UK.4) Data for Ireland is 2005. 5) the range refers to the difference between the highest and lowest sub-group unemployment rate. Average annual changes are presented in yellow. 6) EA without Ireland. 7) Based on fi gures smaller than the Eurostat reliability limit.

78ECB


June 2008

Table 23 Regional mismatch in the euro area

Country1)

Unemployment rate in 2006 (%) Regional mismatch ( range)Change 3) (percentage points)

Min Max Weighted average of subgroups

2006 2) 1984-19954) 1996-2001 2002-2006

Belgium (11) 4.2 17.7 8.3 13.5 4.2 0.3 -4.0 -0.7 4.1 0.8 Germany (41) 5.4 19.7 10.3 14.3 6.1 0.6 6.4 1.1 -5.1 -1.0 Ireland (2) 4.3 4.7 4.4 0.4 -4.5 -0.8 -3.0 -0.5 -1.1 -0.2 Greece (13) 7.2 14.4 9.0 7.2 2.4 0.2 0.6 0.1 -2.5 -0.5 Spain (17) 5) 5.3 13.5 8.6 8.2 3.7 0.4 -6.3 -1.1 -5.8 -1.2 France (22) 6) 6.1 12.9 9.1 6.8 2.4 0.2 -1.4 -0.2 -2.3 -0.5 Italy (21) 2.7 13.6 6.9 11.0 9.0 0.7 2.2 0.4 -13.0 -2.6 Luxembourg (1) 4.7 4.7 4.7 n.a. n.a. n.a. n.a. n.a. n.a. n.a. Netherlands (12) 2.7 5.2 3.9 2.5 -2.4 -0.2 -2.4 -0.4 0.3 0.1 Austria (9) 3.0 8.9 4.8 5.9 n.a. n.a. 0.8 0.1 1.9 0.4 Portugal (7) 3.9 9.5 8.1 5.6 -2.8 -0.3 -2.9 -0.5 1.0 0.2 Slovenia (1) 6.1 6.1 6.1 n.a. n.a. n.a. n.a. n.a. n.a. n.a. Finland (5) 3.5 11.4 7.8 7.9 n.a. n.a. -5.5 -0.9 -4.6 -0.9 Euro area (162) 7) 2.7 19.7 8.4 17.0 2.5 0.3 -6.3 -1.0 -7.3 -1.5 Denmark (1) 4.0 4.0 4.0 n.a. n.a. n.a. n.a. n.a. n.a. n.a.Sweden (8) 6.0 8.5 7.1 2.5 n.a. n.a. -1.2 -0.2 -1.0 -0.2United Kingdom (37) 2.6 9.0 5.4 6.4 -3.5 -0.3 1.8 0.3 0.5 0.1

Sources: EU-LFS (annual, regional data, at NUTS2-level) and NBB calculations. For all tables on mismatch, the fi gures presented are those fulfi lling Eurostat's publications and reliability limits for the LFS data. The weighted average of subgroups may not equal the total unemployment rate due to missing data. In bold are the three best euro area performers who attain (i) a relatively low level and (ii) low range of unemployment rates. Average annual changes are presented in yellow.Notes: 15 to 64 years old.1) Number of regions in 2006 between parentheses.2) No regional mismatch indicator available for Luxembourg, Slovenia and Denmark, as those countries consist of only 1 NUTS2-region. The result for Ireland has to be taken with caution, as this country only consists of 2 NUTS2-regions.3) The developments of the mismatch indicator can be biased by changes in the number of regions considered, or by changes in their territories. This is especially the case in Greece, where the number of regions increased from 9 to 13 in 1988; in Germany, with changes in 1991 (after the reunifi cation), 1996, 2001, 2002 and 2004; and fi nally in Ireland, where in 1998 8 regions were regrouped into 2. As such, the euro area fi gures must also be considered with caution. In the UK, the number of regions increased strongly in 1996.4) 1985-1995 for Germany, 1987-1995 for Spain, Portugal and the euro area, and 1990-1995 for Ireland.5) Ceuta and Melilla are considered to be part of the Andalucia region.6) Excluding the French overseas departments, for which data are only available from 2001 onwards. As those departments have very high unemployment rates, including them would lead to a strong increase of the mismatch indicator over time, which does not provide a correct view.7) The euro area fi gure refl ects the dispersion of unemployment rates within and among the 13 euro area countries. It is not available before 1996.

unemployment rates lower than national total

unemployment rates. These broad patterns also

hold for tertiary level education holders only

and young people aged 15 to 29 years

(see Table 34 and 35 in Annex 3). These latter

specialisations therefore present areas where

the labour demand for skills is high relative to

supply. Furthermore, there are large labour

market imbalances in some subject areas across

countries.130 In general, shortages in the supply

of particular skills may suggest national

education and training systems’ failure to

identify and/or provide the skills demanded by

fi rms. Alternatively, they may indicate that a

subject area is relatively unattractive to workers

– for example due to low rates of return and/or

a lack of relative wage fl exibility.

The broadly stable range in unemployment rates

shows that mismatch by subject specialisation

in the euro area as a whole remained almost

constant over the period considered, again

suggestive of a persistent labour market

misbalance.

For example, many unemployment rates in 2006 are signifi cantly 130

lower than international estimates of the average NAIRU

for the euro area, which tend to cluster around 8% in 2006

(see European Central Bank, 2006).

79ECB


June 2008

5 MATCHING

LABOUR SUPPLY

AND DEMAND5.2.3 REGIONAL MISMATCH 131

Table 23 presents information on regional

differences in unemployment rates across the

euro area countries. Mismatch between the

demand and supply of labour across regions can

have a number of origins, including uneven

regional economic development, concentrations

of the resident population and changes in a

country’s industrial structure. However, if

labour is geographically mobile and wages are

fl exible, such mismatches should be short-lived

as unemployed workers move from one region

to another or regional wages adjust to refl ect

local labour market conditions, and investment

fl ows to regions with cost advantages. In this

respect, persistent regional mismatch can

therefore provide some evidence on wage

infl exibility and/or labour immobility. Indeed,

Table 23 shows that unemployment rates

generally vary considerably across regions of

the euro area. Looking at the 2006 levels, there

are relatively large ranges of unemployment

rates in a number of countries. Whilst regional

mismatch appears limited in the Netherlands

and Ireland132, it is very high in Germany,

Belgium and Italy, suggesting a combination of

low regional mobility and low wage fl exibility

to local conditions133. In a number of these

countries clear regional subdivisions appear.

For example, in Belgium134 and Italy,

unemployment is much higher in the southern

regions135; in Germany unemployment is high in

the eastern part of the country and relatively low

in the south.

Regional disparities as measured by the change

in the range of unemployment rates increased

slightly in the euro area over the 1984-1995

period. During the last decade however136,

regional mismatch in the euro area decreased

strongly; in particular, favourable developments

were recorded in Spain, Italy and Finland.

Since 2001, regional dissimilarities also

diminished strongly in Germany and only

clearly increased in Belgium. The decreases in

the regional variation in unemployment rates

are most likely driven by improvements in the

economic environment. However, they may also

provide an indication of some improvements

in the mobility of labour and regional wage

fl exibility, albeit from a low level.

5.2.4 MISMATCH BY AGE, GENDER AND

NATIONALITY 137

Table 24 shows that the unemployment rate of

young people is very high and higher than that of

other age groups in the euro area. In 2007, youth

unemployment (those aged 15 to 24) stood at

almost 15%, nearly three times the unemployment

rate of both prime-aged workers (aged 25

to 54) and those aged 55 to 64. Furthermore,

unemployment among teenagers aged 15 to

19 reached 19.3% and was higher than among

young adults aged 20 to 24.138 High rates of youth

unemployment are of particular concern since

young people should arguably constitute the

group most interested in building up their human

capital and career and who are most fl exible,

both in terms of their subject specialisation and

geographical mobility. Table 36 in Annex 3

shows that rates of unemployment are particularly

high for young people with lower secondary

education and less, across all euro area countries.

But in some countries rates are also very

high for those with upper secondary and even

tertiary education. High youth unemployment

therefore seems to refl ect a major failure of

education systems in identifying and providing

the appropriate level and area of skill demanded

Prepared by J. De Mulder.131

In addition, the average unemployment rate is also low in the 132

Netherlands and Ireland.

See Vamvakidis (2008).133

In Belgium, part of the regional mismatch may also be due to 134

language barriers between the northern and southern regions,

where different languages are spoken.

In Belgium, the regions Brussels and Wallonia experience 135

relatively higher unemployment. In Italy, unemployment rates

are proportionally high in the whole southern part of the country

(Mezzogiorno) and on the islands of Sicily and Sardinia.

The number of regions considered hardly changed over the 136

period 1996-2006, so the recorded developments for the euro

area are more reliable for this period.


Although for some countries it should be noted that while the 138

unemployment rate for this group is high, the participation rate

is quite low.

80ECB


June 2008

by fi rms. It may also refl ect institutional barriers

that prevent job creation and can cause a (further)

deterioration of human capital, demoralisation

and future labour market diffi culties for the

affected individuals. The pattern of higher rates

of youth unemployment is consistent across all

euro area countries, with particularly high rates

in Belgium, Spain, France, Italy, Luxembourg

and Finland in 2007. Relatively low rates seem

to coincide with countries that have a strong

vocational training system (such as Austria and

Germany – see also Box 8). Unemployment

rates of those aged over 55 were also relatively

high in Germany compared with other euro area

countries.

Whilst the range of unemployment rates

across age groups in the euro area decreased

signifi cantly over the 1983 to 2001 period,

it has increased again since then (mismatch

across the level of education has also increased

for the young, see Table 36 in Annex 3).

It stood at a high level of 12.8 percentage

points in 2007. Overall, this is suggestive of a

persistent and structurally driven labour market

imbalance for the euro area as a whole. The

lack of a positive development in recent years is

worrisome, not least since unemployment rate

differences between age groups are affected by

demographic factors. Following an ageing and

shrinkage of the euro area population and an

associated decrease in the labour supply from

the young (see Chapter 3), these factors might

have been expected to even out the variation in

unemployment rates across age groups to some

extent. In addition, this aggregate fi gure masks

particularly high levels of and rapidly growing

mismatches in some countries in recent years.

Two other important changes in the composition

of labour supply in the euro area over the past

25 years are worth consideration with regard

Table 24 Unemployment rates and mismatch by age groups in the euro area and euro area countries 1)

Country

Unemployment rate in 2007 (%) Mismatch by age (range)Change (p.p) 5)

15-19 20-24 25-54 55-64

Weighted average of subgroups 2007 1984-1995 1996-2001 2002-2007

Belgium 30.2 17.6 6.8 3.8 7.7 26.3 2.9 0.2 -11.2 -1.9 8.2 1.4 Germany 13.5 11.6 7.8 10.1 8.6 5.6 -8.7 -0.7 1.8 0.3 -0.1 0.0 Ireland 2) 13.9 7.4 4.0 2.6 4.6 11.3 -1.3 -0.1 -13.9 6) -2.3 6) 5.2 6) 0.9 6) Greece 24.9 21.5 7.6 3.4 8.2 21.5 8.8 0.7 -0.1 0.0 -9.2 -1.5 Spain 29.1 14.9 6.9 5.6 8.0 23.5 -1.9 -0.2 -13.9 -2.3 -0.5 -0.1 France 28.6 18.6 7.5 6.6 8.7 22.0 -1.3 -0.1 -6.9 -1.2 4.6 0.8 Italy 3) 29.2 16.3 5.0 2.3 5.8 26.9 0.3 0.0 -4.2 -0.7 -3.9 -0.7 Luxembourg n.a. 11.9 6) 3.3 n.a. 3.9 n.a. n.a. n.a. n.a. n.a. n.a. n.a. Netherlands 4) 8.7 4.0 2.5 3.5 3.2 6.2 13.9 -1.2 -7.1 6) -1.2 6) 1.3 6) 0.2 6)

Austria 8.9 7.5 4.2 3.3 4.7 5.7 n.a. n.a 0.2 0.0 2.6 0.4 Portugal 4) 24.8 13.2 7.8 6.8 8.4 17.9 -6.7 6) -0.7 6) -3.2 6) -0.5 6) 9.0 6) 1.5 6) Slovenia 5) 7.4 6) 8.0 4.4 6) 3.1 6) 4.7 5.0 6) n.a. n.a -5.7 6) -0.9 6) -12.8 6) -2.1 6) Finland 33.7 14.6 5.3 5.8 7.8 28.4 n.a. n.a -7.9 -1.3 -3.5 -0.6 Euro area 19.3 13.8 6.6 6.5 7.5 12.8 -7.2 -0.6 -6.9 -1.1 3.1 0.5 Denmark 9.4 5.4 2.7 4.2 3.6 6.7 -9.8 -0.8 0.5 0.1 1.5 0.3 Sweden 6) 37.8 15.3 4.5 4.0 7.0 33.9 n.a. n.a 1.0 0.2 19.6 3.3 United Kingdom 20.7 10.6 3.7 3.3 5.2 17.4 -7.6 -0.6 0.2 0.0 7.2 1.2

Sources: EU-LFS (spring data) and ECB calculations.Notes: For all tables on mismatch, the fi gures presented are those fulfi lling Eurostat's publications and reliability limits for the LFS data. In bold are the three best euro area performers who attain (i) a relatively low level and (ii) low range of unemployment rates. The weighted average of subgroups may not equal the total unemployment rate owing to missing data. 1) 15 to 64 years old. 2) Data start in 1995. 3) Data start in 1986. 4) Data start in 1996. 5) Average annual changes are presented in yellow. The range refers to the difference between the highest and lowest subgroup unemployment rate. 6) Based on fi gures smaller than the Eurostat reliability limit.

81ECB


June 2008

5 MATCHING

LABOUR SUPPLY

AND DEMAND

to mismatch – namely the rapid increase in

the labour participation of women and in

immigration to the euro area (for details of these

developments see Chapter 3).

Table 25 shows that the unemployment rate of

non-nationals was above that of nationals in

2007, particularly so for non-EU nationals, and

across most countries of the euro area.139 The

high rate of unemployment for non-EU nationals

may partly refl ect country-specifi c immigration

experiences with, e.g. asylum seekers and/or the

magnitude of immigrant stocks, and fl ows and

should be compared with the information on the

size of the non-national group presented in

Table 7 and Table 30 in Annex 3. For example,

high rates of unemployment for non-nationals

may refl ect temporary bottlenecks in the

employment of new immigrants, resulting from

increases in immigrant fl ows. However, the

generally relatively high rates of unemployment

for non-nationals may also suggest major

ineffi ciencies in the institutions governing the

integration of non-nationals into the labour

market and/or policies determining immigration

in some countries of the euro area. They could

also highlight the diffi culties that some countries

may face if migration increases, but institutions/

policies do not change. The range of

unemployment rates by nationality stood at a

very high level in many euro area countries,

particularly Belgium, Finland and France in

2007, although trends over the 1996 to 2007

period are suggestive of decreasing labour

market imbalances in the euro area as a whole.

Table 38 in Annex 3 shows rates of unemployment are generally 139

higher for non-nationals across both educational level and age

group. Unfortunately the lack of data prevents a more detailed

breakdown of unemployment rates based on many dimensions

at once.

Table 25 Unemployment rates and mismatch by nationality in the euro area and euro area countries 1)

Country


NationalsOther EU15

Non-EU15

of which 12 NM

of which non-EU27

Weighted average of subgroups 2007

Change (p.p) 7)

1996-2001 2002-2007

Belgium 6.9 9.7 27.1 12.7 30.2 7.7 20.2 -8.0 -1.3 -1.0 -0.2Germany 7.9 9.3 19.2 13.2 20.3 8.6 11.4 -1.7 -0.3 3.7 0.6Ireland 2) 4.4 n.a. n.a. n.a. n.a. 4.4 n.a. n.a. n.a. n.a. n.a.Greece 8.2 n.a. 8.0 8.4 7.9 8.2 n.a. n.a. n.a. n.a. n.a.Spain 7.3 10.7 12.1 11.5 12.3 8.0 4.8 2.9 0.5 -3.0 -0.5France 8.2 7.0 28.0 21.2 8) 24.4 8.3 17.3 -0.5 -0.1 -1.2 -0.2Italy 3) 5.7 n.a. 7.7 7.3 7.8 5.8 n.a. n.a. n.a. n.a. n.a.Luxembourg 3.7 3.3 8) n.a. n.a. n.a. 3.9 n.a. 3.1 8) 0.5 8) n.a. n.a.Netherlands 4) 3.1 n.a. 10.0 n.a. 10.4 3.2 n.a. n.a. n.a. n.a. n.a.Austria 3.9 5.3 12.3 7.2 13.5 4.7 8.4 n.a. n.a. 3.2 0.5Portugal 4) 8.1 n.a. 14.7 n.a. 15.2 8.4 n.a. n.a. n.a. n.a. n.a.Slovenia 5) 4.6 n.a. n.a. n.a. n.a. 4.7 n.a. n.a. n.a. n.a. n.a.Finland 7.6 n.a. 21.0 n.a. 26.4 7.8 13.4 n.a. n.a. n.a. n.a.Euro area 7.0 8.2 14.7 11.0 15.5 7.5 7.7 -0.1 0.0 -1.6 -0.3Denmark 3.4 n.a. 11.9 n.a. 12.2 3.7 8.5 n.a. n.a. n.a. n.a.Sweden 6) 6.7 6.1 19.6 24.5 18.9 7.0 13.5 n.a. n.a. 2.8 0.5United Kingdom 5.0 7.1 8.2 6.0 9.1 5.2 3.2 -1.9 -0.3 -3.5 -0.6

Sources: EU-LFS (spring data) and ECB calculations. For all tables on mismatch, the fi gures presented are those fulfi lling Eurostat's publications and reliability limits for the LFS data. In bold are the three best euro area performers who attain a low level of unemployment rates. The weighted average of subgroups may not equal the total unemployment rate due to missing data. See also Tables A11 and A12 in Appendix 3 for information on unemployment by educational level for non nationals. 1) 15 to 64 years old; the non-national population is separated into non-national EU15 citizens and non-national non-EU15 citizens. For the period 2005-07, this last group is further split into the 12 new member states (which together with the EU15 form the EU27) and non-national non-EU27 citizens. 2) Only data for 1998-2004. 3) Only data for 2005-07. 4) Data start 1999. 5) Data start 2002. 6) Data start 1997. 7) The range refers to the difference between the highest and lowest subgroup unemployment rate. Average annual changes are presented in yellow. 8) Based on fi gures smaller than the Eurostat reliability limit.

82ECB


June 2008

Box 10

THE GENDER PAY GAP AND DISCRIMINATION 1

The promotion of a lifecycle approach to work includes increasing female labour market

participation and reducing gender gaps in pay. The Lisbon Strategy for Growth and Jobs stresses

the need to narrow gender pay gaps as a component of providing the appropriate incentives

for workers to enter the labour market and fostering employment, improved work quality and

productivity, and social cohesion. Eurostat’s structural indicators show that the unadjusted

gender pay gap 2 has decreased only very slightly in the euro area, from 15% in 1995 to 14% in

2005. Furthermore, while many countries in the euro area experienced a signifi cant decrease in

gender pay differentials over this period, in a few euro area countries, the gap between average

male and female pay actually increased 3.

Part of the unadjusted gender pay gap in the euro area may be explained by differences in the

average levels of human capital, such as work experience, education and training, or other

characteristics such as the average age of working men and women. Microeconomic studies

have therefore attempted to estimate to what extent observable characteristics (such as age, work

experience and educational level) explain the unadjusted gender pay gap. Weichselbaumer and

Winter-Ebmer (2005) undertake a meta-analysis of 260 of these studies, fi nding that the average

“explained” component of the unadjusted gap stands at about 30%, the remaining 70% being

“unexplained” - which in this literature is often attributed to discrimination.

Reasons for the persistence in the gender wage gap include gender segregation by sector,

occupation and function. It is still the case that girls more frequently study traditionally female

subjects such as languages and crowd into lower paying sectors and occupations, and part-

time employment, which tend to lead to lower paying careers 4. Furthermore, in areas where

technological change results in increases in relative demand and higher relative wages (such as

IT and engineering – see also Box 9), women are typically underrepresented in both universities

and fi rms. The gender pay gap also tends to increase throughout the professional career, with

the male/female wage gap widening with age, experience and rank. This can be explained fi rst

of all by the fact that women tend to interrupt their professional career more often than men, as

they often take more family and household responsibilities than their partners, and second by the

fact that women’s educational levels were lower in the past. Gender differences in promotion

propensities (the “glass ceiling”) and job-to-job mobility also play a role, having a marked

positive infl uence on wage changes and thus on the development of gender wage gaps.

All in all, weaknesses in the labour market situation of women exist that may reduce their incentives

and opportunity for participating in the labour market and ascending the career paths it offers.

Encouraging girls to invest in education and in a greater variety of fi elds are steps in the right direction,

but will not suffi ce unless contemporaneous measures are undertaken to improve equal opportunities

at work. It is necessary to enhance incentives (and reduce rigidities which hinder fi rms) for paying

wages in line with individual productivity; it is also necessary to continue promoting gender equality

1 Prepared by M.Ward-Warmedinger.

2 The unadjusted gender pay gap is given as the difference between average gross hourly earnings of male paid employees and of

female paid employees as a percentage of average gross hourly earnings of male paid employees. The population consists of all paid

employees aged 16-64 that are ‘at work over 15 hours per week’.

3 For more information and data, see Eurostat’s Structural Indicators.

4 See, for example, European Commission (2002b) for a survey of national evidence on these issues.

83ECB


June 2008

5 MATCHING

LABOUR SUPPLY

AND DEMAND

Table 26 presents the relative unemployment

rates for men and women in the euro area. This

shows a higher unemployment rate for women

relative to men in most euro area countries in

2007, which suggests that also the increase in

female labour supply has not been fully matched

by labour demand (despite the generally

lower average wages of female workers – see

Box 10), resulting in the ineffi cient use of some

of this increased pool of labour supply in the

euro area. This gap between male and female

employment rates has decreased somewhat in

the euro area as a whole and in many of the

euro area countries over the period 1983 to

2007. While higher unemployment rates for

women and non-nationals may refl ect objective

economic reasons and mismatch due to, e.g.

skill acquisition, and subject specialisation,

they may also be due to other factors

(e.g. discrimination).

5.3 THE FUTURE DEMAND FOR LABOUR

It is impossible to fully anticipate future labour

demand needs, and the problem of matching

labour demand and labour supply will remain.

Nevertheless, some components of future labour

demand that have general implications for the

desired composition of labour supply can be

expected.

in male and female-dominated sectors and access to higher functions (such as managerial positions)

to allow women to progress in their careers. Increased accessibility to childcare should allow women

to more easily combine work and family (see Section 4.2). Finally, in some cases improvements in

the effectiveness of legislation against discrimination may also be necessary.

Table 26 Unemployment rates and mismatch by gender in the euro area and euro area countries 1)

Country

Unemployment rate in 2007 (%) Mismatch (range)Male Female Weighted

average of subgroups 2007

Change (p.p) 6)

1984-1995 1996-2001 2002-2007

Belgium 6.7 8.8 7.7 2.0 -4.8 -0.4 -3.6 -0.6 0.7 0.1 Germany 8.5 8.8 8.6 0.3 -0.7 -0.1 -2.6 -0.4 0.3 0.0 Ireland 4.9 4.3 4.6 0.6 -1.4 -0.1 0.2 0.0 0.3 0.0 Greece 5.0 12.8 8.2 7.7 1.5 0.1 1.5 0.2 -1.4 -0.2 Spain 3) 6.1 10.5 8.0 4.4 7.2 0.6 -5.1 -0.8 -3.3 -0.6 France 8.2 9.3 8.7 1.0 -0.3 0.0 -0.6 -0.1 -2.5 -0.4 Italy 4.6 7.5 5.8 2.8 -1.6 -0.1 -1.4 -0.2 -2.8 -0.5 Luxembourg 4.1 3.5 7) 3.9 0.6 7) -0.5 0.0 -1.7 -0.3 -0.0 7) -0.0 7) Netherlands 2.8 3.7 3.2 0.8 -0.3 0.0 -1.8 -0.3 0.1 0.0 Austria 2) 4.4 5.0 4.7 0.7 n.a. 0.0 -0.8 -0.1 0.5 0.1 Portugal 3) 6.9 10.0 8.4 3.1 -3.7 -0.3 0.9 0.1 0.9 0.1 Slovenia 5) 3.7 5.9 4.7 2.2 n.a. 0.0 0.1 0.0 1.6 0.3 Finland 2) 7.5 8.1 7.8 0.6 n.a. 0.0 -0.9 -0.1 -0.2 0.0 Euro area 6.6 8.6 7.5 1.9 0.3 0.0 -1.6 -0.3 -1.0 -0.2 Denmark 3.3 4.1 3.6 0.8 1.8 0.1 -1.8 -0.3 -0.4 -0.1 Sweden 4) 6.5 7.5 7.0 0.9 n.a. 0.0 -1.4 -0.2 0.2 0.0 United Kingdom 5.5 4.9 5.2 0.7 1.0 0.1 -2.2 -0.4 -0.4 -0.1

Sources: EU-LFS (spring data) and ECB calculations. For all tables on mismatch, the fi gures presented are those fulfi lling Eurostat's publications and reliability limits for the LFS data. In bold are the three best euro area performers who attain (i) a relatively low level and (ii) low range of unemployment rates. The weighted average of subgroups may not equal the total unemployment rate due to missing data. 1) 15 to 64 years old. 2) Data start in 1995. 3) Data start in 1986. 4) Data start in 1995.5) Data start in 1996. 6) The range refers to the difference between the highest and lowest subgroup unemployment rate. Average annual changes are presented in yellow. 7) Based on fi gures smaller than the Eurostat reliability limit.

84ECB


June 2008

First, an increasing demand for higher skills

resulting from technological developments can

be expected. While this evolution could help to

ease some of the tensions in countries or regions

with an over-education 140 problem, ceteris paribus it will aggravate the existing mismatches

in others. Furthermore, a change in the

production structure of the economy owing to

technological progress may lead to relative

changes in demand for educational groups.

Second, the consequences of an ageing

population and the rising proportion of the

elderly will affect both the size and the

composition of the euro area labour force. The

shrinking of the total labour force and the ageing

of society will increase demand for certain

subject-specifi c skills such as medical workers

and caregivers in the euro area.

Third, the ongoing process of globalisation may

infl uence the structure of future labour demand.

Changes in the structure of international trade

in goods and services may lead to changes in

labour demand. For example, increased trade

with countries with a large supply of relatively

low-wage, low-skilled labour is likely to induce

a decrease in demand for low-skilled workers

in the production of tradable goods in the euro

area. Similarly, this may increase the demand

for high-skilled workers in the euro area as

emerging markets seek to import capital goods

and technology from abroad. The effects of

demographic change and the increased demand

for services may mean that the demand for

labour in services, including caregivers and

medical workers, will further increase.

These factors mean that the euro area labour

markets must not only be adequately prepared

with an appropriately enhanced quantity

and quality of labour supply, but must

also rise to the task of ensuring a suffi cient

match between this workforce and labour

demand. This may imply, for instance,

greater investment by national authorities in

education systems and greater incentives for

schools, universities and fi rms to identify and

cultivate those skills sought by fi rms. More

general competences will also be needed to

allow individuals/students to adapt fl exibly

to new developments in labour demand.

Globalisation presents a risk to low-skilled

people by increasing the supply of foreign

goods produced using lower wage low-skilled

labour abroad. It is important that wages are

fl exible – to provide clear signals with regard

to which skills are demanded by fi rms and to

reduce wage rents, which privilege ‘insiders’

thereby supporting employment creation.

Furthermore, regulations and institutions

should be reformed to reduce the costs and

ineffi ciencies of the labour marketing process.

Lastly, more coordinated efforts across the

euro area to eliminate the remaining barriers

to geographical mobility (languages, transfer

of pensions, etc.) are needed.

Over-education refers to a situation where workers are undertaking 140

tasks of a lower level than that for which they are qualifi ed.

85ECB


June 2008

ANNEX 1

ANNEX 1 MEASUREMENT ISSUES141

When addressing labour supply, it has to be noted

that offi cial statistics may potentially mis-measure

the true employment and labour market

participation rates. This annex briefl y reviews the

main measurement issues relating to accurately

capturing labour supply. First, the labour force is

equal to the sum of the population at work and

unemployed job-seekers. In the LFS, the ILO-

defi nitions of “at work” 142 and “unemployed” are

used (see also Annex 2). However, some

individuals who do not work but are available for

employment, and thus are a part of actual labour

supply, are not captured by the standard defi nition

of unemployment, which results in an

underestimation of actual labour supply. Box 11

discusses the ILO defi nition of unemployment and

considers how this affects measured employment.

Second, some work is not captured by statistics

on employment because individuals work in the

unoffi cial (or “shadow”) economy. This results

in an underestimate of employment, and to the

extent that these workers are measured as inactive,

an underestimate of actual labour supply, when

supply is measured by the participation rate.

Third, some services that could be provided

through the market or by government are

produced in the household instead. Different

propensities for household production across

countries or over time may infl uence measures

of labour supply. In particular there has been

an increasing trend towards market-based or

government provision of services that have been

traditionally produced within the household (e.g.

childcare). This trend has affected female labour

supply in particular. Freeman and Schettkat

(2005) collect and assess information on the

amount of hours worked in the household across

countries based on time-diaries. They conclude,

for example that there is a more signifi cant

provision of services typically considered as a

part of household production through the market

in the United States than in European countries.

The extent to which household production is

provided via the market is also likely to depend on

the supply of suitable, typically low-skilled and

low-wage labour. Furthermore, part of services

that are “outsourced” from the household may

rather fall into the shadow economy. Generally

as female labour supply increases these effects

may become more important.

Prepared by J. Turunen and M.Ward-Warmedinger.141

Persons who during the reference period performed work for a 142

wage or salary, or for profi t or family gain, in cash or in kind,

for at least one hour. This includes those with a job or enterprise

who are not at work due to temporary absence.

Box 11

ILO DEFINITION OF UNEMPLOYMENT 1

Information on labour market status is computed from data collected by national labour force surveys

in many developed countries. Labour force statistics generally divide the adult population into three

mutually exclusive groups: the employed, the unemployed, and the inactive (i.e. people out of the

labour force). The European Community Labour Force Survey is based on the ILO-defi nition of

“at work” and “unemployed”. Under this defi nition, the employed comprise all persons aged 15

and over 2 who, during some reference period, were engaged in paid employment (including those

working in a family business). The LFS additionally specifi es that individuals are employed if they

engage in paid employment for at least one hour per week. This includes those with a job or enterprise

who are not at work due to temporary absence. In the LFS, people between the ages of 15 and 74 3

are classifi ed as “unemployed” if they meet all of the following requirements: (1) they are without

1 Pepared by E. Viviano.

2 16 and over in Spain, the United Kingdom and Sweden (1995-2001); 15-74 years in Denmark and Sweden (from 2001 onwards).

3 Those aged 16-74 in Spain, Sweden (1995-2000) and the United Kingdom.

86ECB


June 2008

work (work less than one hour per week); (2) they state that they are seeking employment for at

least one hour per week; (3) they are available to start work within the following two weeks; (4) they

sought employment at some time during the previous four weeks (see Eurostat 1996). People neither

employed nor unemployed are considered inactive (and are excluded from the labour force).

People out of the labour force are thus a composite group formed by persons who do not want a job,

persons who are not searching but might take up a job if offered, and persons who are searching for

a job but took their last step to search for one more than four weeks ago. Brandolini et al. (2006)

calculate that, on average in European countries, about a fi fth of all people who declared they were

seeking work in the 1990s were left out of the labour force on the basis of this four-week requirement.

Because of the sheer size of this group – also labelled “potential labour force” or simply “potentials” –

Brandolini et al. investigate the role of the four-week criterion in determining the size of unemployment

and conversely of the potential labour force. They test whether transition probabilities differ among

the unemployed, the potentials, and the other inactive persons in European countries and fi nd that

the (annual) transition probabilities of potentials are always different from those observed for other

people out of the labour force, whereas in some cases they can be considered similar to those of

the unemployed. On this basis, they conclude that in the European labour markets a sizeable “grey

area” exists between the states of unemployment and inactivity. The European labour markets would

be better described by four distinct states (employed, unemployed, potentials, and other inactive

population) than by the three-way characterisation of the ILO guidelines, confi rming the conclusion

reached by Jones and Riddell (1999) for Canada and by Centeno and Fernandes (2004) for Portugal.

The ILO four-week requirement can be viewed as a minimum level of search intensity that job seekers

must show in order to be classifi ed as unemployed: at least one search action – such as sending an

application to a potential employer, visiting an employment agency, or (in Europe) simply looking at

newspaper advertisements – must be undertaken in a four-week period.4 However, this condition may

be exceedingly rigid. From the theoretical standpoint the total effort put into a job search depends on

individual resources, search costs, and expected returns; moreover, it is endogenously determined,

given the labour market conditions. As a consequence, whether this arbitrarily set minimum level

of search intensity is a good criterion for distinguishing between active and less-active job seekers is

ultimately an empirical question. Brandolini et al. (2006) identify the search intensity that separates

the unemployed from the potentials by looking at the Italian data. They proxy search intensity by

the “number of months since the last search action” using data from the Italian labour force survey,

the only EU survey where this information is available. They compare the (quarterly) transition

probabilities of the unemployed with those of a group comprising the most-intensive job seekers

among the potentials. They fi nd that the potentials turn out to be behaviourally indistinguishable

from the unemployed when their last search action occurred “not long before” the ILO four weeks

(up to 11 months for certain groups in the population). Letting the boundary between unemployment

and potential labour force be determined by the data, rather than by the arbitrary four-week criterion,

would raise the Italian unemployment rate in 2000 from 11% to 13%. Four weeks may be too short a

period to identify the unemployed, especially in some demographic groups, like people living in the

Southern part of Italy and older women.

The same exercise, repeated for the years 2004, 2005 and 2006, suggests that, by letting the size

of unemployment be determined by the data, would increase the Italian unemployment rate by

1.5 percentage points (see Bank of Italy, 2006).

4 Alternatively, search intensity may be identifi ed with the probability of applying for a job during a given period or with the number of

applications sent per unit of time (as in Petrongolo and Pissarides, 2001).

87ECB


June 2008

ANNEX 2

ANNEX 2 DATA SOURCES AND DESCRIPTIONS

A2.1 THE LABOUR FORCE SURVEY:

Euro area data presented and used in this paper

are drawn from the European Community

Labour Force Survey, which has been conducted

since 1983. The annual data used consist of the

spring surveys (quarter 1 surveys for France and

Austria, quarter 2 for all other countries) up to

the fi rst quarter of 2007. Eurostat compiles these

data, and a detailed description of the sampling

methods and adjustment procedures can be found

in “The European Union Labour Force Survey

– Methods and Defi nitions, 2001” (http://circa.

europa.eu/irc/dsis/employment/info/data/eu_lfs/

index.htm). The available variables are listed

and described in the “EU Labour Force Survey

database – User guide”. The need to preserve an

international comparability of the data means

that the LFS dataset uses standardised and

widely accepted defi nitions of, e.g. employment

and unemployment, as adopted by the ILO.

These constitute the basis of the Eurostat LFS.

It should be noted that these defi nitions differ

from those adopted by countries in their national

defi nitions of labour market status, where

international comparability is not a necessity.

No data are available for Spain and Portugal

prior to 1986, for Austria and Finland prior to

1995 and for Slovenia prior to 1996.

The LFS is based on the ILO-defi nition of “at

work” and “unemployed”. In the LFS, persons

aged 15 and over 143 are defi ned as “at work” or

employed if during the reference period they

performed work for a wage or salary, for profi t

or family gain (e.g. in a family business), in cash

or in kind, for at least one hour. 144 This includes

those with a job or enterprise who are not at

work due to temporary absence. People aged 15

to 74 145 are classifi ed as “unemployed” if they

meet all of the following requirements: (1) they

are without work (work less than one hour per

week); (2) they state that they are seeking

employment for at least one hour per week; (3)

they are available to start work within the

following two weeks; (4) they sought

employment at some time during the previous

four weeks. People neither employed nor

unemployed are considered inactive (and are

excluded from the labour force). According to

the LFS defi nition, “inactive people” are a group

formed by persons who do not want a job,

persons who are not searching but might take a

job if offered, and persons who are searching

for a job but took their last step to fi nd one more

than four weeks before the interview. Box 11

discusses the implications of these defi nitions

further.

Nationality is interpreted as citizenship.

Citizenship is defi ned according to national

legislation of each country. Education level is

classifi ed according to the International Standard

Classifi cation of Education 1997 (see below).

The expression ‘level successfully completed’

is associated with obtaining a certifi cate or a

diploma, when there is a certifi cation. In cases

where there is no certifi cation, successful

completion must be associated with full

attendance. When determining the highest level,

both general and vocational education/training

is taken into consideration.

Data on the hours of work measure the number of

hours usually worked per week. This covers all

hours including extra hours, either paid or unpaid,

which the person normally works, but excludes

travel time between home and workplace and

time taken for the main meal break (usually at

lunchtime). Persons who usually work from

home are asked to include the number of hours

they usually work there. Apprentices, trainees and

other persons learning a job are asked to exclude

any time spent at college or in other special

training centres. Some persons, particularly self-

employed persons and family workers, may not

have usual hours, in the sense that their hours

vary considerably from week to week or month

to month. If a respondent is unable to provide a

fi gure for usual working hours for this reason, the

average of hours actually worked per week over

16 and over in Spain, the United Kingdom and Sweden (1995-143

2001); 15-74 years in Denmark and Sweden (from 2001 onwards).

This group therefore includes employees, employers employing 144

one or more employees, the self-employed and family workers.

Those aged 16-74 in Spain, Sweden (1995-2000) and the United 145

Kingdom.

88ECB


June 2008

the past four weeks is used as a measure of usual

hours. The distinction between full-time and part-

time work is based on a spontaneous response

by the respondent (except in the Netherlands,

where part-time is determined if the usual hours

are fewer than 35 hours and full-time if the usual

hours are 35 hours or more, and in Sweden where

this criterion is applied to the self-employed).

It has to be noted that methodological changes

due to, e.g. census revisions, changes in concepts

and a change from the use of annual to quarterly

or more frequent surveys and data, have resulted

in breaks in the LFS survey in many euro area

countries. Details of these breaks can be found

at: http://circa.europa.eu/irc/dsis/employment/

info/data/eu_lfs/F_LFS_COMPARABILITY.

htm. Comparability of developments over time

in this report is optimised through the use of the

spring surveys for all countries over the relevant

time period considered.

DATA EXTRAPOLATION

In the case of Germany, data prior to 1991

have been obtained on the basis of the

developments in West Germany. The euro

area aggregate refers to the 13 countries that

formed the euro area in 2007. Data prior to

1996 have been obtained on the basis of the

growth rate of the largest aggregate available

(i.e. 13 countries since 1996, 12 countries in

1995, 10 countries from 1986 to 1994 and

8 countries for 1983 to 1985).

DEFINITION OF EDUCATIONAL LEVELS IN THE

EU-LFS

In the EU-LFS, the ISCED 1997 classifi cation

is used. Respondents are considered as low,

medium or high skilled if their most advanced

diploma is respectively ISCED 0, 1 or 2;

ISCED 3 or 4; and ISCED 5 or 6.

ISCED 1 - PRE-PRIMARY EDUCATION

Programs at level 0, (pre-primary) defi ned as

the initial stage of organised instruction are

designed primarily to introduce very young

children to a school-type environment, i.e. to

provide a bridge between the home and a school-

based atmosphere. Upon completion of these

programs, children continue their education at

level 1 (primary education).

ISCED 2 - PRIMARY EDUCATION OR FIRST STAGE

OF BASIC EDUCATION

Programmes at level 1 are normally designed on a

unit or project basis to give students a sound basic

education in reading, writing and mathematics

along with an elementary understanding of other

subjects such as history, geography, natural

science, social science, art and music. In some

cases religious instruction is featured. The core

at this level consists of education provided for

children, the customary or legal age of entrance

being not younger than fi ve years or older than

seven years. This level covers, in principle, six

years of full-time schooling.

ISCED 3 - LOWER SECONDARY EDUCATION OR

SECOND STAGE OF BASIC EDUCATION

The contents of education at this stage are

typically designed to complete the provision of

basic education which began at ISCED level 1.

In many, if not most countries, the educational

aim is to lay the foundation for lifelong learning

and human development. The programmes at

this level are usually on a more subject-oriented

pattern using more specialised teachers and more

often several teachers conducting classes in their

fi eld of specialisation. The full implementation

of basic skills occurs at this level. The end

of this level often coincides with the end of

compulsory schooling, where it exists.

ISCED 4 - (UPPER) SECONDARY EDUCATION

This level of education typically begins at the

end of full-time compulsory education for those

countries that have a system of compulsory

education. More specialisation may be observed

at this level than at ISCED level 2 and often

teachers need to be more qualifi ed or specialised

than for ISCED level 2. The entrance age to this

level is typically 15 to 16 years. The educational

programmes included at this level typically

require the completion of some 9 years of full-

time education (since the beginning of level 1)

89ECB


June 2008

ANNEX 2

for admission, or a combination of education

and vocational or technical experience.

ISCED 3A: Programmes designed to provide

direct access to ISCED 5A;

ISCED 3B: Programmes designed to provide

direct access to ISCED 5B;

ISCED 3C: Programmes not designed to lead to

ISCED 5A or 5B.

ISCED 5 - POST-SECONDARY NON TERTIARY

EDUCATION

ISCED 4 captures programmes that straddle

the boundary between upper secondary and

post-secondary education from an international

point of view, even though they might clearly

be considered as upper secondary or post-

secondary programmes in a national context.

These programmes can, considering their

content, not be regarded as tertiary programmes.

They are often not signifi cantly more advanced

than programmes at ISCED 3 but they serve to

broaden the knowledge of participants who have

already completed a programme at level 3.

Typical examples are programmes designed

to prepare students for studies at level 5 who,

although they have completed ISCED level 3,

did not follow a curriculum which would allow

entry to level 5, i.e. pre-degree foundation

courses or short vocational programmes. Second

cycle programmes can be included as well.

ISCED 4A: See text for ISCED 3

ISCED 4B: See text for ISCED 3

ISCED 4C: See text for ISCED 3

LEVEL 6 - FIRST STAGE OF TERTIARY EDUCATION

(NOT LEADING DIRECTLY TO AN ADVANCED

RESEARCH QUALIFICATION)

This level consists of tertiary programmes

with a more advanced educational content than

those offered at levels 3 and 4. Entry to these

programmes normally requires the successful

completion of ISCED level 3A or 3B or a

similar qualifi cation at ISCED level 4A. They

do not confer an advanced research qualifi cation

(ISCED 6). These programmes must have a

cumulative duration of at least two years.

ISCED 5A: Programmes that are largely

theoretically based and are intended to provide

suffi cient qualifi cations for gaining entry into

advanced research programmes and professions

with high skills requirements.

ISCED 5B: Programmes that are practically

oriented/occupationally specifi c and are mainly

designed for participants to acquire the practical

skills and know-how needed for employment

in a particular occupation, trade or class of

occupations or trades, the successful completion

of which usually provides the participants with

a labour-market relevant qualifi cation

ISCED 7 - SECOND STAGE OF TERTIARY

EDUCATION (LEADING TO AN ADVANCED

RESEARCH QUALIFICATION)

This level is reserved for tertiary programmes

which confer an advanced research qualifi cation.

The programmes are therefore devoted to

advanced study and original research and not

based on course-work only. They typically

require the submission of a thesis or dissertation

of publishable quality which is the product of

original research and represents a signifi cant

contribution to knowledge. They prepare

graduates for faculty posts in institutions

offering ISCED 5A programmes, as well as

research posts in government, industry, etc.

A2.2 US DATA

Total population data stems from the AMECO

database. Working age population, employment

and unemployment are from the US Bureau

of Labour Statistics (BLS). Each month, the

BLS analyses and publishes statistics on the

labour force, employment, and unemployment,

classifi ed by a variety of demographic, social,

90ECB


June 2008

and economic characteristics. These statistics

are derived from the Current Population Survey

(CPS), which is conducted by the Census

Bureau for the BLS. This monthly survey of the

population uses a sample of households that is

designed to represent the civilian noninstitutional

population of the United States.

The data used in Table 17 are based on

the publication Social Security Programs

Throughout the World: Europe, 2006. This

information is compiled by the US social security

administration and the International Social

Security Association (ISSA). The publication is

available through: http://www.ssa.gov/policy/

docs/progdesc/ssptw/2006-2007/europe/index.

html. This publication contains an overview

of the different social security programs in the

world, including parental leave structures.

A2.3 PISA

The OECD’s Programme for International

Student Assessment (PISA) is an internationally

standardised assessment that was jointly

developed by participating countries and

administered to 15-year-olds in schools. The

assessment conducted by PISA in 2006 covers the

domains of reading, mathematical and scientifi c

literacy, with the major focus on scientifi c

literacy. Paper-and-pencil tests are used, with

assessments lasting a total of two hours for each

student. Test items are a mixture of multiple-

choice items and questions requiring students

to construct their own responses. Students

also respond to a background questionnaire,

and additional supporting information is

gathered from the school authorities. Fifty-six

countries and regions, including all 30 OECD

member countries, took part in the PISA 2006

assessment. The assessment takes place every

three years.

A2.4 OECD DATA ON YEARLY HOURS WORKED

The OECD uses information from labour force

surveys: for European countries this is the

EU-LFS.

Annual hours are estimated by annualising usual

weekly hours and then by adding or subtracting

all “unusual” events that occurred during the

year, such as additional overtime work, public

holidays, sick, maternity, paid and other leave.

The number of weeks corresponding to each

of the events can be estimated by using the

differences between actual and usual weekly

hours reported in the survey and extrapolating

them over the year. However, for events that

are not randomly distributed over the year, such

as paid leave and public holidays, statutory

information is used. For the European countries,

information concerning collectively agreed paid

leave and public holidays are taken from the

Working Time Developments of the European

Industrial Relations Observatory (EIRO).

The estimation of average annual hours worked

is done separately for full-time employees, part-

time employees and self-employed persons.

Average annual hours for total employment are

then calculated as a weighted average of the

three categories.

91ECB


June 2008

ANNEX 3

ANNEX 3 DETAILED TABLES AND CHARTS

Chart 11 Employment rates in the euro area and the United States by gender

(percentages)

Euro area

United States

Males Females

55

65

75

85

55

65

75

85

20041983 1986 1989 1992 1995 1998 200140

50

60

70

40

50

60

70

20041983 1986 1989 1992 1995 1998 2001

Sources: Eurostat and ECB calculations.Note: 15 to 64 year olds.

Table 27 Participation and employment rates for five-year age groups in the euro area

Males Females1983 1995 2001 2007 1983 1995 2001 2007

Participation rates Total 80.4 76.4 77.1 78.4 46.9 54.3 57.9 63.2

15-19 37.9 24.8 26.4 25.4 31.6 19.5 20.9 20.4

20-24 80.3 67.6 68.0 68.1 66.0 59.8 58.2 59.7

25-29 93.2 88.6 88.1 88.7 63.8 72.0 73.9 77.4

30-34 97.5 95.0 94.7 94.4 58.9 70.1 73.6 77.9

35-39 97.7 96.0 95.6 95.5 55.5 69.6 73.3 77.7

40-44 97.2 95.6 95.1 94.8 51.9 69.3 73.6 78.1

45-49 95.5 93.6 93.5 93.3 49.4 62.9 70.4 76.4

50-54 90.6 87.2 87.8 89.9 41.8 53.9 60.2 69.7

55-59 73.6 66.1 66.7 71.6 30.4 35.8 41.0 51.7

60-64 40.3 28.2 30.0 36.3 13.2 11.3 13.3 20.5

Employment ratesTotal 73.5 69.1 71.8 73.2 40.9 46.6 52.3 57.8

15-19 28.3 19.3 22.4 20.8 21.0 13.9 16.9 16.1

20-24 64.9 53.8 58.8 59.3 51.5 45.1 48.8 50.8

25-29 83.9 77.8 80.3 81.3 54.5 59.8 65.1 69.7

30-34 91.3 86.9 88.8 88.7 53.3 60.5 66.4 71.3

35-39 93.0 89.5 90.7 90.7 51.0 61.3 67.1 71.7

40-44 92.8 89.3 90.6 90.0 48.3 62.2 67.8 72.8

45-49 91.2 87.6 89.0 88.8 46.6 57.1 65.5 71.3

50-54 86.1 81.5 83.3 85.2 39.3 48.7 55.8 64.8

55-59 69.7 60.2 61.6 67.0 28.7 32.1 37.1 47.9

60-64 38.0 26.7 28.2 34.2 12.8 10.9 12.7 19.3

Sources: EU-LFS (spring data), ECB and NBB calculations.

92ECB


June 2008

Table 28 Participation rates (and employment rates in brackets) by country and subgroup

a) Subdivision according to gender

Average annual change (percentage points) Level (%) Trend developments Recent developments

1984 – 1995 1) 1996 – 2007 2) 1996 – 2001 3) 2002 – 2007 2007

MalesBelgium -0.3(-0.2) 0.1(0.1) 0.1(0.3) 0.1(-0.1) 73.2(68.2)

Germany -0.2(-0.1) 0.1(0.0) -0.1(-0.2) 0.4(0.3) 81.4(74.4)

Ireland -0.6(-0.4) 0.4(0.9) 0.5(1.6) 0.3(0.2) 81.2(77.2)

Greece -0.4(-0.4) 0.1(0.2) 0.0(-0.1) 0.3(0.5) 78.9(74.9)

Spain -0.4(-0.1) 0.5(1.2) 0.5(1.8) 0.6(0.7) 81.6(76.6)

France -0.5(-0.7) 0.0(0.1) 0.0(0.4) -0.1(-0.3) 74.3(68.2)

Italy -0.5(-0.7) 0.1(0.4) 0.1(0.3) 0.1(0.5) 74.5(71.1)

Luxembourg -0.4(-0.4) 0.0(-0.2) 0.0(0.1) -0.1(-0.4) 75.5(72.4)

Netherlands 0.2(0.5) 0.4(0.6) 0.7(1.3) 0.1(-0.1) 84.7(82.3)

Austria n.a.(n.a.) 0.0(-0.1) -0.3(-0.3) 0.2(0.1) 80.2(76.7)

Portugal -0.7(-0.6) 0.2(0.2) 0.5(0.9) 0.0(-0.6) 79.0(73.6)

Slovenia n.a.(n.a.) 0.4(0.6) 0.3(0.5) 0.6(0.8) 76.0(73.2)

Finland n.a.(n.a.) 0.4(1.0) 0.8(1.7) 0.0(0.3) 79.3(73.4)

Euro area -0.3(-0.4) 0.2(0.3) 0.1(0.4) 0.2(0.2) 78.4(73.2)

Denmark 0.1(0.4) -0.1(0.0) -0.4(-0.1) 0.1(0.2) 84.0(81.3)

Sweden n.a.(n.a.) 0.2(0.4) 0.1(0.8) 0.3(0.1) 82.0(76.7)

United Kingdom -0.1(0.0) -0.1(0.2) -0.2(0.5) -0.1(-0.1) 81.6(77.1)

FemalesBelgium 0.6(0.8) 0.7(0.8) 0.5(0.9) 0.9(0.7) 60.2(54.9)

Germany 0.7(0.7) 0.7(0.7) 0.4(0.6) 1.0(0.8) 69.8(63.7)

Ireland 0.6(0.7) 1.3(1.6) 1.5(2.1) 1.2(1.1) 63.1(60.3)

Greece 0.4(0.3) 0.9(0.8) 0.9(0.6) 0.9(1.1) 55.1(48.1)

Spain 1.3(0.7) 1.3(1.9) 0.8(1.8) 1.8(2.0) 61.2(54.8)

France 0.4(0.1) 0.4(0.6) 0.3(0.6) 0.5(0.6) 65.1(59.0)

Italy 0.2(0.1) 0.7(0.9) 0.8(0.9) 0.6(1.0) 50.6(46.8)

Luxembourg 0.3(0.3) 0.9(0.9) 1.3(1.4) 0.6(0.4) 55.4(53.5)

Netherlands 1.5(1.6) 1.2(1.4) 1.4(2.0) 0.9(0.7) 72.2(69.6)

Austria n.a.(n.a.) 0.4(0.4) 0.0(0.1) 0.8(0.7) 67.2(63.8)

Portugal 0.6(0.8) 0.8(0.6) 0.9(1.2) 0.7(0.1) 68.6(61.7)

Slovenia n.a.(n.a.) 0.5(0.5) 0.2(0.2) 0.8(0.8) 67.2(63.3)

Finland n.a.(n.a.) 0.5(0.9) 0.9(1.4) 0.1(0.4) 75.3(69.2)

Euro area 0.6(0.5) 0.7(0.9) 0.6(0.9) 0.9(0.9) 63.2(57.8)

Denmark 0.1(0.2) 0.3(0.5) 0.3(0.7) 0.2(0.3) 76.4(73.3)

Sweden n.a.(n.a.) 0.2(0.2) 0.0(0.5) 0.3(-0.1) 77.7(71.9)

United Kingdom 0.8(0.8) 0.2(0.3) 0.3(0.6) 0.1(0.0) 68.6(65.2)

Sources: EU-LFS (spring data), ECB and NBB calculations. Notes: 15 to 64 years old. The fi gures corresponding to employment rates are given in brackets. 1) 1987-1995 for Spain and Portugal. 2) 1997-2007 for Slovenia. 3) 1997-2001 for Slovenia.

93ECB


June 2008

ANNEX 3

Table 28 Participation rates (and employment rates in brackets) by country and subgroup (cnt’d)

b) Subdivision according to age

15-24 years old


1984-1995 1) 1996-2007 2) 1996-2001 3) 2002-2007 2007

Belgium -0.8 (-0.6) -0.1 (0.0) 0.0 (0.3) -0.1 (-0.3) 33.1 (26.8)

Germany -0.2 (0.2) -0.2 (-0.4) -0.3 (-0.3) -0.1 (-0.5) 49.7 (43.7)

Ireland -1.3 (-0.9) 0.7 (1.0) 0.8 (1.7) 0.5 (0.3) 53.1 (48.4)

Greece -0.5 (-0.5) -0.5 (-0.2) 0.0 (0.0) -0.9 (-0.3) 31.0 (24.2)

Spain -0.6 (-0.1) 0.5 (1.2) 0.1 (1.6) 0.9 (0.9) 47.8 (39.1)

France -1.5 (-1.4) 0.1 (0.3) 0.0 (0.6) 0.3 (0.0) 37.3 (29.6)

Italy -0.8 (-0.7) -0.6 (0.0) -0.4 (0.1) -0.9 (-0.1) 31.0 (25.3)

Luxembourg -1.6 (-1.5) -1.3 (-1.3) -1.1 (-1.0) -1.4 (-1.7) 26.0 (22.1)

Netherlands 1.1 (1.3) 0.9 (1.2) 1.9 (2.6) -0.1 (-0.3) 73.0 (68.6)

Austria n.a. (n.a.) -0.2 (-0.3) -1.2 (-1.1) 0.8 (0.5) 59.2 (54.5)

Portugal -2.0 (-1.4) -0.2 (-0.1) 0.6 (1.0) -0.9 (-1.3) 40.9 (34.7)

Slovenia n.a. (n.a.) -0.2 (0.2) -1.3 (-1.0) 0.7 (1.2) 40.4 (37.2)

Finland n.a. (n.a.) 1.0 (1.6) 2.2 (2.8) -0.1 (0.4) 62.1 (48.6)

Euro area -0.7 (-0.5) 0.0 (0.3) 0.0 (0.6) 0.0 (0.0) 44.0 (37.3)

Denmark 0.7 (1.1) 0.0 (0.1) -1.0 (-0.7) 0.9 (1.0) 72.6 (67.4)

Sweden n.a. (n.a.) 0.8 (0.5) 1.2 (1.6) 0.5 (-0.7) 55.1 (42.1)

United Kingdom -0.1 (0.2) -0.4 (-0.2) -0.3 (0.3) -0.4 (-0.8) 59.2 (50.8)

25-24 years oldBelgium 0.5 (0.5) 0.4 (0.5) 0.1 (0.5) 0.7 (0.5) 85.1 (79.3)

Germany 0.4 (0.3) 0.4 (0.3) 0.4 (0.4) 0.4 (0.2) 87.7 (80.8)

Ireland 0.5 (0.6) 0.8 (1.2) 1.0 (1.9) 0.6 (0.4) 82.2 (78.9)

Greece 0.5 (0.4) 0.7 (0.6) 0.6 (0.4) 0.7 (0.8) 82.0 (75.7)

Spain 1.1 (0.5) 0.7 (1.5) 0.3 (1.7) 1.1 (1.3) 82.8 (77.1)

France 0.4 (0.0) 0.1 (0.3) 0.0 (0.4) 0.3 (0.3) 87.7 (81.1)

Italy 0.1 (-0.1) 0.5 (0.7) 0.5 (0.6) 0.4 (0.8) 77.5 (73.6)

Luxembourg 0.4 (0.4) 0.8 (0.7) 1.0 (1.1) 0.5 (0.2) 82.8 (80.1)

Netherlands 0.9 (1.0) 0.7 (0.9) 0.8 (1.4) 0.6 (0.4) 87.6 (85.4)

Austria n.a. (n.a.) 0.3 (0.3) 0.3 (0.4) 0.2 (0.1) 86.7 (83.0)

Portugal 0.7 (0.7) 0.4 (0.2) 0.3 (0.7) 0.4 (-0.2) 87.7 (80.9)

Slovenia n.a. (n.a.) 0.3 (0.4) 0.2 (0.4) 0.3 (0.3) 89.9 (85.9)

Finland n.a. (n.a.) 0.2 (0.9) 0.5 (1.4) 0.0 (0.3) 88.3 (83.7)

Euro area 0.5 (0.2) 0.4 (0.6) 0.3 (0.7) 0.5 (0.5) 84.6 (79.1)

Denmark -0.2 (0.0) 0.1 (0.4) 0.1 (0.5) 0.2 (0.3) 88.5 (86.1)

Sweden n.a. (n.a.) 0.0 (0.3) -0.3 (0.3) 0.4 (0.3) 90.3 (86.3)

United Kingdom 0.4 (0.4) 0.1 (0.3) 0.0 (0.5) 0.1 (0.1) 84.5 (81.3)

Sources: EU-LFS (spring data), ECB and NBB calculations. Corresponding fi gures for employment rates between brackets.1) 1987-1995 for Spain and Portugal. 2) 1997-2007 for Slovenia. 3) 1997-2001 for Slovenia.

94ECB


June 2008


b) Subdivision according to age

55-64 years old


1984-1995 1) 1996-2007 2) 1996-2001 3) 2002-2007 2007

Belgium -0.5 (-0.5) 0.9 (0.9) 0.3 (0.3) 1.5 (1.4) 35.2 (33.8)

Germany 0.2 (-0.1) 1.3 (1.2) 0.0 (0.0) 2.5 (2.4) 57.9 (52.0)

Ireland -0.4 (-0.4) 1.0 (1.2) 0.8 (1.2) 1.3 (1.2) 55.5 (54.0)

Greece -0.5 (-0.5) 0.1 (0.1) -0.4 (-0.4) 0.6 (0.7) 43.6 (42.1)

Spain -0.6 (-0.5) 0.9 (1.1) 0.9 (1.2) 1.0 (0.9) 47.4 (44.8)

France -0.6 (-0.6) 0.8 (0.7) 0.2 (0.2) 1.3 (1.2) 40.5 (37.8)

Italy -0.5 (-0.5) 0.5 (0.5) -0.1 (-0.1) 1.1 (1.2) 34.8 (34.0)

Luxembourg -0.1 (-0.1) 0.9 (0.9) 0.1 (0.1) 1.6 (1.6) 34.5 (34.3)

Netherlands -0.2 (-0.1) 1.9 (1.8) 1.7 (1.7) 2.2 (2.0) 52.8 (51.0)

Austria n.a. (n.a.) 0.7 (0.7) -0.2 (-0.3) 1.6 (1.6) 38.3 (37.1)

Portugal 0.2 (0.1) 0.5 (0.4) 0.8 (0.9) 0.3 (-0.1) 54.0 (50.3)

Slovenia n.a. (n.a.) 1.4 (1.4) 0.8 (0.7) 1.9 (1.9) 36.0 (34.9)

Finland n.a. (n.a.) 1.6 (1.7) 1.7 (1.9) 1.5 (1.6) 58.8 (55.4)

Euro area -0.3 (-0.4) 0.9 (0.9) 0.2 (0.3) 1.5 (1.5) 46.4 (43.4)

Denmark 0.0 (-0.1) 0.6 (0.8) 0.9 (1.2) 0.4 (0.4) 61.3 (58.7)

Sweden n.a. (n.a.) 0.5 (0.7) 0.4 (0.7) 0.6 (0.6) 72.9 (69.9)

United Kingdom -0.1 (0.0) 0.7 (0.8) 0.4 (0.8) 0.9 (0.9) 59.3 (57.4)

Sources: EU-LFS (spring data), ECB and NBB calculations. Corresponding fi gures for employment rates between brackets.1) 1987-1995 for Spain and Portugal. 2) 1997-2007 for Slovenia. 3) 1997-2001 for Slovenia.

c) Subdivision according to education level


1996-2007 1) 1996-2001 2) 2002-2007 2007

LowBelgium 0.2 (0.2) 0.0 (0.3) 0.4 (0.2) 55.9 (49.3)

Germany 0.9 (0.6) 0.7 (0.7) 1.1 (0.4) 66.4 (54.7)

Ireland 0.4 (0.8) 0.3 (1.3) 0.4 (0.3) 62.8 (58.7)

Greece 0.4 (0.3) 0.3 (0.1) 0.5 (0.6) 64.5 (59.9)

Spain 0.6 (1.2) 0.4 (1.3) 0.9 (1.0) 66.5 (60.9)

France 0.1 (0.3) 0.2 (0.4) 0.1 (0.1) 64.6 (57.4)

Italy 0.2 (0.4) 0.2 (0.2) 0.3 (0.5) 56.1 (52.7)

Luxembourg 0.4 (0.3) 0.5 (0.6) 0.2 (0.0) 60.7 (58.1)

Netherlands 0.6 (0.8) 0.8 (1.5) 0.5 (0.3) 64.1 (61.4)

Austria 0.0 (-0.1) -0.8 (-0.9) 0.9 (0.7) 61.6 (56.4)

Portugal 0.5 (0.3) 0.7 (1.0) 0.3 (-0.3) 77.5 (71.2)

Slovenia 0.5 (0.5) 1.1 (1.0) 0.0 (0.1) 61.5 (57.2)

Finland 0.0 (0.5) 0.2 (0.8) -0.1 (0.2) 65.1 (59.3)

Euro area 0.4 (0.6) 0.3 (0.7) 0.5 (0.5) 63.5 (57.6)

Denmark 0.5 (0.9) 0.3 (0.8) 0.8 (0.9) 70.3 (67.5)

Sweden -1.3 (-0.9) -2.3 (-1.5) -0.2 (-0.4) 71.5 (66.6)

United Kingdom -0.4 (-0.1) -0.6 (-0.1) -0.1 (-0.1) 68.6 (64.4)

Sources: EU-LFS (spring data), ECB and NBB calculations. Notes: 25 to 64 years old. EU-LFS data concerning education level are only available from 1992 onwards. The fi gures corresponding to employment rates are given in brackets. 1) 1997-2007 for the Netherlands and Slovenia. 2) 1997-2001 for the Netherlands and Slovenia.

95ECB


June 2008

ANNEX 3


c) Subdivision according to education level


1996-2007 1) 1996-2001 2) 2002-2007 2007

MediumBelgium 0.0 (0.1) 0.0 (0.4) 0.1 (-0.2) 78.8 (73.9)

Germany 0.5 (0.5) 0.2 (0.2) 0.8 (0.8) 81.6 (74.8)

Ireland 0.7 (0.9) 1.2 (1.7) 0.2 (0.0) 80.2 (77.5)

Greece 0.5 (0.5) 0.7 (0.5) 0.3 (0.6) 75.7 (69.6)

Spain 0.2 (1.0) -0.3 (1.2) 0.7 (0.8) 81.9 (76.6)

France -0.2 (0.0) -0.2 (0.1) -0.2 (-0.2) 80.3 (74.9)

Italy 0.1 (0.3) 0.1 (0.2) 0.1 (0.5) 78.1 (75.1)

Luxembourg 0.4 (0.4) 0.7 (0.8) 0.1 (0.0) 75.9 (74.1)

Netherlands 0.4 (0.6) 0.7 (1.2) 0.2 (0.1) 82.6 (80.4)

Austria 0.0 (0.0) -0.2 (-0.1) 0.3 (0.2) 78.5 (75.7)

Portugal 0.2 (0.2) 0.4 (0.9) 0.0 (-0.5) 85.4 (79.7)

Slovenia 0.0 (0.0) -0.3 (-0.1) 0.1 (0.1) 78.6 (75.1)

Finland 0.1 (0.7) 0.4 (1.3) -0.2 (0.2) 81.8 (77.0)

Euro area 0.2 (0.4) 0.1 (0.3) 0.4 (0.4) 80.4 (75.3)

Denmark 0.1 (0.4) 0.1 (0.5) 0.2 (0.3) 84.2 (82.2)

Sweden -0.3 (0.0) -0.9 (-0.2) 0.2 (0.2) 87.1 (83.3)

United Kingdom 0.0 (0.2) 0.1 (0.7) -0.1 (-0.2) 84.2 (81.1)

HighBelgium 0.1 (0.1) 0.0 (0.1) 0.2 (0.1) 87.8 (84.9)

Germany 0.2 (0.3) -0.1 (0.0) 0.5 (0.5) 89.8 (86.5)

Ireland 0.2 (0.3) 0.3 (0.7) 0.1 (0.0) 89.0 (87.0)

Greece 0.3 (0.3) 0.2 (0.0) 0.3 (0.5) 88.4 (83.3)

Spain 0.1 (0.8) -0.2 (1.0) 0.4 (0.6) 88.8 (84.6)

France -0.1 (0.1) 0.0 (0.3) -0.1 (-0.1) 87.2 (83.0)

Italy -0.3 (-0.1) -0.2 (0.1) -0.4 (-0.2) 84.1 (80.6)

Luxembourg 0.3 (0.1) 0.5 (0.5) 0.0 (-0.3) 86.5 (84.0)

Netherlands 0.3 (0.4) 0.3 (0.8) 0.2 (0.2) 89.6 (88.1)

Austria -0.2 (-0.2) -0.4 (-0.3) 0.1 (0.0) 88.9 (86.7)

Portugal 0.0 (-0.2) 0.0 (0.2) 0.0 (-0.6) 92.5 (86.9)

Slovenia 0.3 (0.3) -0.1 (0.0) 0.6 (0.5) 91.6 (89.1)

Finland 0.0 (0.3) 0.2 (0.7) -0.2 (-0.1) 88.2 (85.2)

Euro area 0.0 (0.2) -0.1 (0.3) 0.2 (0.2) 88.3 (84.7)

Denmark -0.1 (0.1) -0.2 (0.1) 0.0 (0.1) 90.2 (87.6)

Sweden -0.1 (-0.1) -0.7 (-0.4) 0.4 (0.2) 91.9 (88.6)

United Kingdom 0.0 (0.1) 0.0 (0.3) 0.0 (0.0) 89.8 (88.0)

Sources: EU-LFS (spring data), ECB and NBB calculations. Notes: 25 to 64 years old. EU-LFS data concerning education level are only available from 1992 onwards. The fi gures corresponding to employment rates are given in brackets. 1) 1997-2007 for the Netherlands and Slovenia. 2) 1997-2001 for the Netherlands and Slovenia.

96ECB


June 2008


d) Subdivision according to nationality

Nationals

Average annual change (percentage points) Level (%)Trend

developments Recent developments1996-2007 1996-2001 2002-2007 1) 2007

Belgium 0.4 (0.4) 0.3 (0.5) 0.4 (0.3) 67.0 (62.4)

Germany 0.5 (0.4) 0.2 (0.2) 0.8 (0.7) 76.7 (70.6)

Ireland n.a. (n.a.) n.a. (n.a.) 0.6 (0.5) 71.4 (68.2)

Greece 0.6 (0.6) 0.5 (0.3) 0.6 (0.8) 66.6 (61.1)

Spain 0.8 (1.5) 0.6 (1.8) 1.1 (1.3) 70.5 (65.3)

France 0.2 (0.3) 0.2 (0.5) 0.2 (0.1) 70.0 (64.2)

Italy n.a. (n.a.) n.a. (n.a.) n.a. (n.a.) 61.9 (58.4)

Luxembourg 0.4 (0.3) 0.5 (0.6) 0.3 (0.0) 61.7 (59.4)

Netherlands n.a. (n.a.) n.a. (n.a.) 0.4 (0.3) 79.1 (76.7)

Austria 0.3 (0.3) -0.1 (-0.1) 0.7 (0.6) 74.2 (71.3)

Portugal n.a. (n.a.) n.a. (n.a.) 0.3 (-0.2) 73.4 (67.5)

Slovenia n.a. (n.a.) n.a. (n.a.) 0.6 (0.8) 71.7 (68.4)

Finland 0.4 (1.0) 0.8 (1.6) 0.0 (0.4) 77.4 (71.6)

Euro area 0.3 (0.5) 0.4 (0.8) 0.2 (0.3) 70.9 (65.9)

Denmark 0.1 (0.3) 0.0 (0.4) 0.2 (0.3) 81.2 (78.5)

Sweden n.a. (n.a.) n.a. (n.a.) 0.3 (0.0) 80.4 (75.1)

United Kingdom 0.0 (0.2) 0.0 (0.5) 0.0 (-0.1) 75.2 (71.4)

Other EU15 citizensBelgium 0.5 (0.9) 0.2 (1.0) 0.8 (0.7) 69.1 (62.4)

Germany 0.2 (0.2) 0.1 (0.3) 0.3 (0.0) 76.7 (69.6)

Ireland n.a. (n.a.) n.a. (n.a.) n.a. (n.a.) n.a. (n.a.)

Greece 0.6 (-0.6) 0.4 (-0.6) -1.6 (-0.6) 53.6 (48.8)

Spain 0.7 (1.1) 0.5 (1.8) 1.0 (0.4) 70.2 (62.7)

France 0.1 (0.3) -0.1 (0.3) 0.4 (0.4) 72.8 (67.8)


Luxembourg 0.4 (0.4) 0.7 (0.9) 0.1 (-0.1) 71.6 (69.3)


Austria 0.8 (0.5) 0.4 (0.0) 1.1 (1.0) 77.2 (73.1)

Portugal n.a. (n.a.) n.a. (n.a.) 1.7 (1.3) 74.6 (69.3)

Slovenia n.a. (n.a.) n.a. (n.a.) n.a. (n.a.) n.a. (n.a.)

Finland 1.5 (2.2) 2.7 (4.9) 0.3 (-0.4) 85.1 (77.8)

Euro area 0.2 (0.3) 0.0 (0.5) 0.4 (0.2) 73.7 (67.6)

Denmark 0.1 (0.3) 1.4 (1.1) -1.2 (-0.5) 76.6 (73.8)

Sweden n.a. (n.a.) n.a. (n.a.) 0.6 (0.3) 75.4 (70.8)

United Kingdom 0.4 (0.6) 0.1 (1.0) 0.7 (0.3) 76.9 (71.5)

Sources: EU-LFS (spring data), ECB and NBB calculations. Notes: 15 to 64 years old. EU-LFS data concerning nationality are only available from 1995 onwards. The fi gures corresponding to employment rates are given in brackets. 1) 2003-07 for Slovenia.

97ECB


June 2008

ANNEX 3


d) Subdivision according to nationality

Non EU15 citizens

Average annual change (percentage points) Level (%)Trend

developments Recent developments1996-2007 1996-2001 2002-2007 1) 2007

Belgium 1.1 (1.1) 0.0 (0.8) 2.1 (1.5) 55.7 (40.6)

Germany 0.0 (-0.1) -0.3 (0.0) 0.3 (-0.2) 63.8 (51.5)

Ireland n.a. (n.a.) n.a. (n.a.) n.a. (n.a.) n.a. (n.a.)

Greece 0.1 (0.5) 0.3 (0.7) 0.0 (0.4) 73.8 (67.9)

Spain 0.7 (1.4) 1.2 (2.2) 0.2 (0.5) 79.7 (70.0)

France 0.1 (0.4) 0.2 (0.6) 0.1 (0.2) 59.3 (44.9)


Luxembourg 0.9 (0.5) 1.5 (1.2) 0.2 (-0.2) 64.9 (57.7)


Austria -0.7 (-1.0) -0.3 (-0.5) -1.2 (-1.5) 67.8 (59.5)

Portugal n.a. (n.a.) n.a. (n.a.) 1.3 (0.4) 82.8 (70.6)

Slovenia n.a. (n.a.) n.a. (n.a.) 2.1 (1.3) 65.7 (59.7)

Finland 0.7 (0.9) 1.6 (1.0) -0.2 (0.9) 67.8 (53.6)

Euro area 0.6 (0.8) 0.2 (0.5) 1.1 (1.1) 69.6 (59.3)

Denmark 0.2 (0.9) -0.9 (0.7) 1.4 (1.2) 61.1 (53.9) Sweden n.a. (n.a.) n.a. (n.a.) 0.3 (-0.2) 65.1 (52.3)

United Kingdom 0.8 (1.2) -0.1 (0.6) 1.7 (1.9) 71.0 (65.2)

Non EU15 citizens: level (%) in 2007

Citizens of the 12 new EU member states

Non EU27 citizens

Belgium 71.1 (62.1) 53.2 (37.1)

Germany 73.3 (63.7) 62.4 (49.7)

Ireland n.a. (n.a.) n.a. (n.a.)

Greece 71.7 (65.6) 74.2 (68.4)

Spain 82.4 (72.9) 79.0 (69.2)

France 68.0 (53.6) 58.9 (44.5)

Italy 76.7 (71.1) 72.2 (66.6)

Luxembourg 63.4 (61.4) 65.4 (56.4)

Netherlands 72.3 (67.7) 56.3 (50.4)

Austria 75.1 (69.6) 66.3 (57.3)

Portugal 72.9 (67.8) 83.5 (70.8)

Slovenia n.a. (n.a.) 65.1 (59.0)

Finland 80.0 (74.4) 63.9 (47.0)

Euro area 77.1 (68.7) 68.3 (57.7)

Denmark 76.3 (71.8) 60.3 (52.9)

Sweden 74.1 (56.0) 63.9 (51.9)

United Kingdom 84.7 (79.6) 66.7 (60.7)

Sources: EU-LFS (spring data), ECB and NBB calculations. Notes: 15 to 64 years old. EU-LFS data concerning nationality are only available from 1995 onwards. The fi gures corresponding to employment rates are given in brackets. 1) 2003-07 for Slovenia.

98ECB


June 2008

Table 29 Developments in the percentage of nationals and non-nationals in the working age population, participation and employment, by country

% of working age population

Country Nationals Other EU15 citizens Non EU15 citizens

average annual change level average annual change level average annual change level1996-2001 2002-2007 2007 1996-2001 2002-2007 2007 1996-2001 2002-2007 2007

Belgium -0.1 0.1 90.9 0.1 -0.1 5.3 0.0 0.1 3.8

Germany -0.1 -0.1 89.6 0.0 0.0 2.8 0.1 0.1 7.6

Ireland n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a.

Spain -0.5 -1.5 87.1 0.1 0.1 1.5 0.4 1.4 11.4

Greece -0.4 -0.4 93.8 0.0 0.0 0.3 0.4 0.4 6.0

France 0.0 0.2 94.1 0.0 0.0 2.1 0.0 -0.1 3.8

Italy n.a. n.a. 94.2 n.a. n.a. 0.3 n.a. n.a. 5.5

Luxembourg -0.8 -0.3 57.9 0.5 0.3 37.7 0.3 0.0 4.4

Netherlands n.a. 0.1 95.7 n.a. 0.0 1.6 n.a. -0.1 2.7

Austria -0.1 -0.3 88.7 0.1 0.1 2.2 0.0 0.2 9.1

Portugal n.a. -0.3 96.3 n.a. 0.0 0.4 n.a. 0.3 3.3

Slovenia n.a. 0.0 99.3 n.a. n.a. n.a. n.a. 0.0 0.7Finland -0.1 -0.1 98.0 0.0 0.0 0.4 0.1 0.1 1.6

Euro area -0.1 -0.2 91.9 0.0 -0.1 1.8 0.1 0.3 6.3

Denmark -0.2 -0.3 94.6 0.0 0.0 1.1 0.2 0.3 4.4

Sweden n.a. 0.1 95.0 n.a. 0.0 2.1 n.a. 0.0 2.9

United Kindgom -0.2 -0.4 92.2 0.0 0.0 1.8 0.2 0.4 6.0

Participation rates Belgium 0.3 0.4 67.0 0.2 0.8 69.1 0.0 2.1 55.7

Germany 0.2 0.8 76.7 0.1 0.3 76.7 -0.3 0.3 63.8


Spain 0.6 1.1 70.5 0.5 1.0 70.2 1.2 0.2 79.7

Greece 0.5 0.6 66.6 0.4 -1.6 53.6 0.3 0.0 73.8

France 0.2 0.2 70.0 -0.1 0.4 72.8 0.2 0.1 59.3


Luxembourg 0.5 0.3 61.7 0.7 0.1 71.6 1.5 0.2 64.9

Netherlands n.a. 0.4 79.1 n.a. 0.2 78.0 n.a. 0.9 57.6

Austria -0.1 0.7 74.2 0.4 1.1 77.2 -0.3 -1.2 67.8

Portugal n.a. 0.3 73.4 n.a. 1.7 74.6 n.a. 1.3 82.8

Slovenia n.a. 0.5 71.7 n.a. n.a. n.a n.a. 1.8 65.7Finland 0.8 0.0 77.4 2.7 0.3 85.1 1.6 -0.2 67.8

Euro area 0.4 0.2 70.9 0.0 0.4 73.7 0.2 1.1 69.6

Denmark 0.0 0.2 81.2 1.4 -1.2 76.6 -0.9 1.4 61.1

Sweden n.a. 0.3 80.4 n.a. 0.6 75.4 n.a. 0.3 65.1

United Kindgom 0.0 0.0 75.2 0.1 0.7 76.9 -0.1 1.7 71.0

Source: EU-LFS (spring data).

Notes: IE: data available for 1998-2004 only. IT: for 2005-07 only. NL and PT: data start 1999 SI: data start 2002 SE: data start 1997.

The non-national population is separated into non-national EU15 citizens and non-national non-EU15 citizens. Numbers in italics are

based on fi gures smaller than the Eurostat reliability limit.

Chart 12 Annual and weekly hours worked in the EU15 countries and the United States in 2005

30 32 34 36 38 40 42 44

2,100

2,000

1,900

1,800

1,700

1,600

1,500

1,400

1,300

2,100

2,000

1,900

1,800

1,700

1,600

1,500

1,400

1,300

NL

DK

DE

AT

UK

SE FR

LU

BEEA

FI

IT

IE

PTES

USGR

x-axis: usual weekly hours worked

y-axis: annual hours worked

Sources: EU-LFS (spring data) and OECD.

99ECB


June 2008

ANNEX 3

Table 29 Developments in the percentage of nationals and non-nationals in the working age population, participation and employment, by country (cnt’d)

% of working age population

Country Nationals Other EU15 citizens Non EU15 citizens

average annual change level average annual change level average annual change level1996-2001 2002-2007 2007 1996-2001 2002-2007 2007 1996-2001 2002-2007 2007

Employment rates 2007

Belgium 0.5 0.3 62.4 1.0 0.7 62.4 0.8 1.5 40.6

Germany 0.2 0.7 70.6 0.3 0.0 69.6 0.0 -0.2 51.5


Spain 1.8 1.3 65.3 1.8 0.4 62.7 2.2 0.5 70.0

Greece 0.3 0.8 61.1 -0.6 -0.6 48.8 0.7 0.4 67.9

France 0.5 0.1 64.2 0.3 0.4 67.8 0.6 0.2 44.9


Luxembourg 0.6 0.0 59.4 0.9 -0.1 69.3 1.2 -0.2 57.7

Netherlands n.a. 0.3 76.7 n.a. 0.2 75.8 n.a. 0.4 51.8

Austria -0.1 0.6 71.3 0.0 1.0 73.1 -0.5 -1.5 59.5

Portugal n.a. -0.2 67.5 n.a. 1.3 69.3 n.a. 0.4 70.6

Slovenia n.a. 0.7 68.4 n.a. n.a. n.a. n.a. 1.0 59.7Finland 1.6 0.4 71.6 4.9 -0.4 77.8 1.0 0.9 53.6

Euro area 0.8 0.3 65.9 0.5 0.2 67.6 0.5 1.1 59.3

Denmark 0.4 0.3 78.5 1.1 -0.5 73.8 0.7 1.2 53.9

Sweden n.a. 0.0 75.1 n.a. 0.3 70.8 n.a. -0.2 52.3

United Kindgom 0.5 -0.1 71.4 1.0 0.3 71.5 0.6 1.9 65.2

Source: EU-LFS (spring data).

Notes: IE: data available for 1998-2004 only. IT: for 2005-07 only. NL and PT: data start 1999 SI: data start 2002 SE: data start 1997.

The non-national population is separated into non-national EU15 citizens and non-national non-EU15 citizens. Numbers in italics are

based on fi gures smaller than the Eurostat reliability limit.

Table 30 Educational Attainment of 25-54 year olds

(% of total population)

1992 1999 2007 Country Low Medium High Low Medium High Low Medium High

Belgium 44.5 31.7 23.7 37.6 33.2 29.2 27.6 38.4 34.0

Germany 16.5 61.1 22.4 17.4 58.6 24.0 14.2 61.0 24.8

Ireland 54.1 27.7 18.2 40.6 37.5 21.9 27.2 37.5 35.3

Greece 56.8 28.5 14.7 42.7 37.9 19.3 34.2 41.5 24.2

Spain 71.6 13.5 14.9 58.5 17.5 24.1 44.5 23.7 31.8

France n.a. n.a. n.a. 34.6 42.4 22.9 27.2 43.1 29.6

Italy 62.3 30.1 7.6 51.5 38.1 10.4 42.8 42.5 14.7

Luxembourg 62.4 24.3 13.3 34.5 45.8 19.7 30.6 39.3 30.1

Netherlands n.a. n.a. n.a. 32.2 44.0 23.8 23.3 44.5 32.1

Austria n.a. n.a. n.a. 21.4 63.6 15.0 17.8 64.0 18.2

Portugal 76.8 11.2 12.0 77.9 12.1 10.0 68.7 16.1 15.3

Slovenia n.a. n.a. n.a. 22.3 61.3 16.3 14.9 60.3 24.8

Finland n.a. n.a. n.a. 22.6 43.7 33.8 14.6 47.2 38.2

Euro area 45.2 38.9 15.9 37.6 41.8 20.6 30.5 44.2 25.4

Denmark 21.4 57.8 20.8 18.0 53.7 28.3 21.9 43.9 34.2

Sweden n.a. n.a. n.a. 19.3 50.2 30.5 11.9 55.0 33.1

United Kingdom 48.8 31.4 19.8 35.3 36.4 28.3 25.6 41.2 33.2

Source: EU – LFS (spring data).

100ECB


June 2008

Table 31 Educational Attainment of Population (% of cohort population by highest level of education attained)

LOW EDUCATION: This Table shows the percentage of a cohort population with a low level of education,

e.g. Belgium 20-24 = 25.6 => that 25.8 % of the 20-24 year old population in Belgium has a low level of education

1992 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64Belgium 25.6 33.6 38.9 41.6 46.3 53.4 61.0 68.3 74.4

Germany 17.6 12.7 13.8 14.1 16.2 18.6 23.9 30.7 35.7

Ireland 32.5 41.8 46.3 51.9 60.2 63.3 68.2 73.2 77.6

Greece 28.8 37.4 45.1 52.4 62.2 68.1 75.4 80.2 84.1

Spain 47.3 54.3 63.1 71.4 79.3 85.0 87.5 90.9 91.9

France n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a.

Italy 44.9 52.1 54.4 57.0 63.9 72.2 78.4 84.3 86.4

Luxembourg 54.4 55.2 59.7 61.9 64.0 68.2 69.5 76.9 77.3

Netherlands n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a.

Austria n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a.

Portugal 65.0 65.3 70.2 75.1 80.6 84.2 88.0 91.2 92.1

Slovenia n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a.

Finland n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a.

Euro area 34.8 35.9 39.4 42.7 48.3 53.7 54.5 63.5 68.3

Denmark 21.3 12.8 15.2 19.3 24.1 24.7 35.6 44.3 48.3

Sweden n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a.

United Kingdom 42.9 46.4 46.8 46.2 47.7 51.4 56.7 61.3 58.5

1999 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64Belgium 23.8 22.5 28.8 35.1 41.9 45.9 53.3 59.6 68.8

Germany 25.4 16.7 15.4 16.5 16.8 18.4 21.7 25.0 32.3

Ireland 18.0 24.9 31.0 37.5 43.4 53.2 60.0 63.0 70.8

Greece 21.4 26.5 32.9 39.0 46.3 53.5 63.3 73.2 79.0

Spain 34.8 42.7 49.1 55.3 62.7 71.4 79.0 84.0 88.9

France 20.0 22.3 26.4 33.0 38.4 42.3 46.7 54.8 63.9

Italy 33.7 40.4 47.3 48.9 50.9 58.4 67.3 74.6 81.1

Luxembourg 28.8 32.3 31.1 32.7 33.9 39.8 40.6 48.7 57.4

Netherlands 27.7 24.4 26.7 29.7 32.9 38.8 42.6 48.0 51.5

Austria 15.3 15.0 16.7 18.7 23.2 29.0 29.5 35.6 47.8

Portugal 59.9 64.6 74.7 78.7 80.0 85.0 87.9 91.3 93.5

Slovenia 14.2 12.4 18.2 18.9 26.5 27.7 31.6 36.9 46.1

Finland 13.2 15.5 13.5 15.5 21.0 27.7 38.9 48.7 59.1

Euro area 28.4 29.3 32.0 35.1 38.7 43.4 50.2 53.3 60.7

Denmark 26.8 10.6 14.8 20.2 20.5 20.2 22.0 25.2 36.4

Sweden 13.7 12.7 12.5 17.1 21.2 23.1 28.6 34.4 41.3

United Kingdom 24.7 30.7 34.8 35.5 34.6 35.7 41.0 48.2 43.8

2007 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64Belgium 17.8 17.6 20.4 23.1 27.8 33.7 41.5 45.0 53.8

Germany 28.1 14.8 15.0 13.6 13.5 13.8 15.2 18.8 20.2

Ireland 13.5 14.3 20.3 24.7 31.6 35.8 45.1 55.9 60.9

Greece 17.6 23.4 25.9 29.4 35.6 42.5 50.9 59.1 68.0

Spain 39.4 34.6 36.2 41.5 46.1 53.8 60.5 68.3 75.9

France 18.4 15.7 18.8 24.2 28.8 34.5 40.6 45.1 49.5

Italy 24.0 28.0 35.3 42.0 46.9 50.4 53.8 61.8 72.1

Luxembourg 29.7 18.0 24.1 29.3 35.8 35.4 38.8 46.9 48.8

Netherlands 24.4 16.4 17.8 21.4 24.3 27.2 31.2 36.9 42.0

Austria 16.1 13.8 13.5 16.0 17.5 21.1 25.0 29.0 29.5

Portugal 47.1 51.2 60.1 69.1 75.7 77.2 81.4 85.1 88.8

Slovenia 9.1 4.5 10.0 14.4 15.7 18.7 26.2 27.5 29.2

Finland 18.2 10.5 9.8 14.8 12.4 16.0 22.5 30.7 40.3

Euro area 26.1 22.5 26.1 28.8 31.1 34.9 39.1 44.8 51.0

Denmark 28.7 15.0 14.7 18.4 23.2 28.2 30.9 31.5 39.2

Sweden 13.0 10.3 8.0 9.0 11.1 14.7 18.4 22.4 29.4

United Kingdom 21.7 19.6 20.4 26.6 27.5 29.0 29.3 34.1 31.1

101ECB


June 2008

ANNEX 3

Table 31 Educational Attainment of Population (% of cohort population by highest level of education attained) (cnt’d)

HIGH EDUCATION: This Table shows the percentage of a cohort population with a high level of education.

1992 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64Belgium 16.2 29.0 26.6 24.6 22.7 19.7 16.2 11.0 8.2

Germany 5.1 16.3 24.1 26.0 24.9 24.2 20.2 17.0 14.6

Ireland 18.1 22.3 20.0 18.8 17.0 15.7 13.3 10.7 8.9

Greece 7.4 20.8 17.7 16.8 13.6 10.9 8.3 6.0 4.7

Spain 11.8 22.3 18.6 15.0 11.8 9.1 7.6 5.7 4.9

France n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a.

Italy 1.1 5.8 8.6 10.2 9.2 6.9 5.2 4.0 4.0

Luxembourg 9.1 14.4 12.9 13.6 14.6 13.2 10.4 6.7 5.8

Netherlands n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a.

Austria n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a.

Portugal 3.9 13.7 15.1 15.0 11.3 8.7 7.4 5.5 4.8

Slovenia n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a.

Finland n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a.

Euro area 6.0 15.0 17.9 18.3 16.4 14.5 12.8 9.8 8.2

Denmark 0.6 11.6 23.1 25.8 24.7 21.3 19.3 12.6 10.6

Sweden n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a.

United Kingdom 12.7 20.8 21.3 21.4 20.5 18.2 15.5 14.1 13.5

1999 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64Belgium 15.4 37.2 32.0 30.5 27.2 25.3 22.4 18.4 13.0

Germany 4.2 17.6 24.5 25.2 26.0 25.8 23.8 21.7 17.6

Ireland 20.9 30.8 24.7 21.3 20.3 17.0 14.8 14.0 11.0

Greece 6.0 21.6 25.4 20.8 18.1 16.2 12.0 9.2 6.5

Spain 22.6 35.2 29.3 25.4 19.5 16.6 12.7 10.0 6.8

France 25.2 33.6 26.2 21.7 20.0 18.7 16.9 13.4 9.7

Italy 1.4 9.1 10.8 11.0 11.9 10.8 8.8 7.0 4.8

Luxembourg 15.9 20.6 21.7 16.9 18.4 21.2 19.3 15.6 8.1

Netherlands 7.9 24.9 25.4 23.7 25.9 23.2 19.4 18.3 15.4

Austria 5.2 14.8 16.3 16.4 14.9 13.8 12.9 12.4 9.3

Portugal 5.2 13.7 11.1 9.2 9.9 8.4 6.6 5.1 3.8

Slovenia 3.3 19.5 15.7 19.4 14.2 14.3 14.9 11.7 11.1

Finland 9.3 35.4 39.2 35.7 35.1 29.9 28.6 24.0 17.1

Euro area 11.7 23.0 22.6 21.2 20.4 19.0 16.4 14.6 11.2

Denmark 3.5 26.1 31.1 28.8 29.0 27.4 27.2 21.3 16.2

Sweden 20.1 31.9 31.8 31.1 29.1 31.8 27.7 22.7 18.3

United Kingdom 22.0 30.6 27.9 27.8 29.1 29.3 24.9 21.0 21.4

2007 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64Belgium 18.1 38.4 39.3 37.7 34.0 29.9 25.8 23.9 20.1

Germany 3.3 18.8 26.2 25.6 25.7 25.4 26.0 23.2 22.1

Ireland 25.3 45.1 41.9 37.8 30.5 26.9 23.1 18.4 16.2

Greece 9.6 28.0 26.8 25.3 25.1 20.7 18.8 15.2 10.9

Spain 20.1 38.3 39.6 34.4 29.9 24.3 20.1 17.4 14.0

France 26.2 42.8 40.7 31.4 24.6 20.9 18.7 17.4 15.6

Italy 7.4 19.1 18.5 16.0 12.1 10.8 11.6 11.1 7.4

Luxembourg 10.4 40.1 36.3 32.3 23.1 22.7 28.5 19.0 20.2

Netherlands 14.5 36.7 36.4 31.7 29.4 30.4 29.6 27.5 23.4

Austria 3.1 16.4 21.6 18.8 17.8 17.3 17.0 13.9 12.9

Portugal 7.1 22.0 19.7 14.6 12.5 11.1 10.5 7.8 6.3

Slovenia 2.4 32.1 31.6 24.4 20.5 22.3 17.9 15.7 15.8

Finland 1.0 26.3 46.3 42.8 41.9 38.5 34.0 28.5 27.0

Euro area 12.8 29.5 30.8 26.6 23.6 21.5 20.5 18.3 15.7

Denmark 4.6 36.9 41.9 35.3 32.0 30.1 29.3 24.4 21.2

Sweden 10.1 39.0 40.6 32.9 28.9 29.1 28.9 27.8 24.5

United Kingdom 21.0 36.4 38.1 32.8 31.9 30.5 30.3 26.4 23.8

102ECB


June 2008

Table 32 Educational attainment expressed in average number of years in formal education 2004

25-to-64-year-old populationMales Females

Total Males Females 25-34 35-44 45-54 54-64 25-34 35-44 45-54 54-64

Belgium 11.3 11.4 11.4 12.4 11.7 11.1 10.3 12.8 11.9 10.7 9.5

Germany 13.4 13.7 13.2 13.6 13.8 13.8 13.7 13.5 13.4 13.2 12.5

Ireland 13.0 12.9 13.1 14.0 13.4 12.3 11.2 14.5 13.6 12.5 11.4

Greece 10.9 11.0 10.7 11.9 11.7 10.9 9.4 12.6 11.7 10.0 8.2

Spain 10.6 10.6 10.6 11.9 11.2 10.1 8.9 12.5 11.4 9.7 8.0

France 11.6 11.7 11.4 12.8 12.1 11.3 10.3 13.1 12.0 10.7 9.6

Italy 10.1 10.2 10.0 11.2 10.5 10.0 8.7 11.7 10.7 9.5 7.6

Luxembourg 13.3 13.6 13.0 14.2 13.5 13.5 13.1 14.1 13.3 12.6 11.6

Netherlands 11.2 11.4 11.1 12.0 11.5 11.3 10.6 12.5 11.4 10.5 9.8

Austria 12.0 12.3 11.7 12.4 12.4 12.2 12.0 12.3 12.0 11.4 10.8

Portugal 8.5 8.3 8.7 9.3 8.4 7.8 7.3 10.3 8.8 7.9 7.2

Slovenia n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a.

Finland 11.2 10.9 11.4 12.5 12.3 10.5 8.5 13.5 13.0 11.2 8.5

Euro area 11.4 11.5 11.4 12.4 11.9 11.2 10.3 12.8 11.9 10.8 9.6

Denmark 13.4 13.5 13.3 13.6 13.6 13.4 13.6 13.6 13.3 13.3 13.0

Sweden 12.6 12.4 12.8 13.1 12.7 12.2 11.3 13.6 13.0 12.7 11.8

United Kingdom 12.6 12.7 12.4 13.0 12.6 12.7 12.4 12.9 12.4 12.3 12.0

Source: OECD “Education at a Glance, 2006” (Table A1.5). Note: Euro area in this table is calculated as a simple average of the available Euro area countries.

Chart 13 Correlation between annual expenditure per student and aggregate Pisa scores

GRPT

ESFR

IT

LU

ATBEIEDE

NL

FI

1,350

1,400

1,450

1,500

1,550

1,600

1,650

1,700

1,350

1,400

1,450

1,500

1,550

1,600

1,650

1,700

12,0002,000 4,000 6,000 8,000 10,000

y-axis: aggregate 2006 Pisa scc

x-axis: annual expenditure per primary student

IEDE

BE

AT

FRESPT

ITGR

LU

NL

FI

1,350

1,400

1,450

1,500

1,550

1,600

1,650

1,700

1,350

1,400

1,450

1,500

1,550

1,600

1,650

1,700

16,0002,000 4,000 6,000 8,000 10,000 12,000 14,000

y-axis: aggregate 2006 Pisa score

x-axis: annual expenditure per secondary student

Source: OECD (2007d) and ECB calculations.

103ECB


June 2008

ANNEX 3

Table 33 Private internal rates of return for an individual obtaining a university-level degree, ISCED 5/6 (2003)

Rate of return when the individual immediately aquires the next higher

level of education

Rate of return when the individual, at age 40, begins the next higher level of education in full time studies, and the

individual bears 1):

Direct costs of foregone earningsNo direct costs but foregone earnings

Males % Females % Males % Females % Males % Females %

Belgium 10.7 15.2 20.0 28.2 21.1 32.2

Denmark 8.3 8.1 12.4 10.2 12.5 10.5

Finland 16.7 16.0 16.2 13.2 16.4 13.4

Sweden 8.9 8.2 10.4 8.2 10.8 8.7

United Kingdom 16.8 19.6 11.4 14.9 12.5 16.8

United States 14.3 13.1 12.9 9.7 15.1 13.0

Source: OECD, “Education at a Glance, 2007”. 1) Private internal rate of return: additions to after-tax earnings that result from higher education net of the additional private costs of education attainment (private expenditures and foregone earnings). Living expenses are excluded from these private expenses. Direct costs are costs of tuition as reported by the national authorities. Foregone earnings are net of taxes.

Table 34 Unemployment rates and mismatch by type of education (tertiary education only) in the euro area and the euro area countries 1)

Unemployment rate in 2006 (%)mismatch

(range)

Country General

Teacher training

and educa-

tion

Humani-ties

langu ages and

art

Social science

business, law

Science math

compu-ting

Engin eering

manufac-turing

Agri-culture

Health and

welfareSer-

vices

Weighted average

of sub groups 2) 2006

Change (p.p.) 5)

2004-2006 3)

Belgium 13.1 2.3 6.2 4.4 4.2 3.5 3.1 2.2 6.1 3.8 4.0 -3.1 -1.6Germany n.a. 3.9 4.3 4.2 5.0 4.9 3.5 3.4 3.6 4.4 1.6 -1.9 -0.6Ireland 4) 2.1 1.6 4.0 1.6 2.6 2.0 0.3 1.2 2.1 2.0 3.7 n.a. n.a Greece n.a. 3.9 8.1 7.6 5.2 4.8 7.5 7.0 4.0 6.3 4.2 -1.0 -0.3Spain n.a. 4.6 7.2 6.2 7.2 4.1 4.8 4.5 6.3 5.6 3.2 -1.4 -0.5France n.a. 3.2 8.8 6.7 6.1 5.8 3.7 2.9 9.5 6 6.7 -0.9 -0.3Italy 3) n.a. 4.6 5.8 6.2 4.5 4.2 5.2 1.6 3.9 4.8 4.6 -0.6 -0.2Luxembourg n.a. 0.6 3.9 2.9 5.5 2.4 6.2 2.4 0.0 2.8 6.2 -0.2 -0.1Netherlands n.a. 1.5 3.8 2.1 2.9 2.4 1.6 2.1 2.6 2.5 2.4 -1.1 -0.4Austria n.a. 2.0 4.3 3.2 2.6 2.1 1.1 1.5 2.8 2.5 3.3 -2.3 -1.2Portugal n.a. 4.8 8.3 5.3 5.8 5.9 1.7 1.8 10.9 5.4 9.1 4.7 2.3Slovenia n.a. 1.4 7.1 3.9 1.8 1.9 2.4 2.5 4.3 3 5.7 -2.0 -0.7Finland n.a. 1.5 6.6 4.0 7.0 2.6 3.7 1.8 2.3 3.3 5.5 0.6 0.2Euro area 6) 2.2 3.5 6.5 5.4 5.6 4.6 3.9 3.1 5.4 4.9 3.4 -0.9 -0.3Denmark n.a. 2.1 6.8 3.7 4.5 2.6 1.9 2.4 3.1 3.2 4.9 -8.0 -2.7 Sweden n.a. 2.6 7.9 5.3 6.6 4.9 3.7 2.2 2.8 4.2 5.7 -2.3 -1.1 United

Kingdom n.a. 1.4 3.0 2.3 2.9 2.0 1.2 1.3 4.6 2.2 3.4 0.4 0.2

Sources: EU-LFS (spring data) and ECB calculations. Notes: For all tables on mismatch, the fi gures presented are those fulfi lling Eurostat's publications and reliability limits for the LFS data. 1) 25 to 64 years old. 2) Differences in the total unemployment rate compared to table 4.1.a are due to missing data. 3) Starting 2004 for AT, BE, IE, PT, SE, UK. 4) Data for Ireland is 2005. 5) Average annual changes are presented in yellow. 6) EA is calculated without IE. The category "general" is excluded from this analysis due to questionable data quality.

104ECB


June 2008

Table 35 Unemployment rates and mismatch by type of education (15-29 year-olds only) in the euro area and the euro area countries 1)


Country General

Teacher training

and educa-

tion

Humani-ties

langu-ages and

art

Social science

business, law

Science math

compu-ting

Engine-ering

manufac-turing

Agri-culture

Health and

welfareSer -

vices

Weighted average of

subgroups 2) 2006

Change (p.p.) 5)

2004-2006 3)

Belgium 15.2 7.6 18.6 13.2 11.7 8.8 6.9 9.6 14.8 14.6 11.8 -5.6 -2.8Germany 5.4 5.0 10.7 10.4 5.7 11.8 13.2 6.7 11.9 11.8 8.2 -3.2 -1.1Ireland 4) 5.2 2.4 5.6 2.3 5.3 3.8 1.3 3.4 3.0 5.9 4.3 n.a. n.a.Greece 17.0 25.5 24.7 20.3 22.4 14.3 27.7 27.4 18.5 17.8 13.4 1.7 0.6Spain 16.2 11.5 12.4 10.4 14.4 7.8 14.2 11.0 13.3 14.1 8.4 -3.6 -1.2France 19.0 0.0 17.4 14.5 8.8 10.1 10.1 9.0 16.3 16.2 19.0 4.7 1.6Italy 24.4 17.0 18.3 15.8 19.8 11.0 12.4 13.5 16.5 15.8 13.3 -15.3 -5.1Luxembourg 15.4 6.1 11.0 7.5 17.4 5.8 0.0 2.9 6.0 10.4 17.4 6.9 2.3Netherlands 7.0 3.3 7.1 3.7 5.2 3.5 5.4 3.8 3.3 6.4 3.9 -2.4 -0.8Austria 6.8 4.2 6.9 6.1 2.9 4.6 5.2 2.3 8.3 7.5 6.0 0.9 0.5Portugal 12.3 19.0 12.9 12.7 15.6 9.7 7.0 13.1 13.3 13.3 12.1 2.7 1.3Slovenia 11.1 5.1 19.8 13.2 6.6 9.2 17.0 10.0 11.6 11.9 14.6 1.9 0.6Finland 18.0 0.0 18.2 7.1 15.8 8.9 12.4 4.8 9.5 18.4 18.2 -0.2 -0.1Euro area 6) 11.8 9.8 15.4 12.1 12.3 10.0 11.1 8.5 12.8 13.6 6.9 -1.9 -0.6Denmark 7.7 3.3 6.3 6.1 6.3 3.8 4.2 5.0 5.7 6.7 4.4 -8.1 -2.7Sweden 12.8 7.8 13.7 10.4 7.1 9.9 12.3 10.5 9.1 15.4 6.6 -3.5 -1.7United

Kingdom 9.7 3.6 5.5 5.9 8.5 5.1 5.6 5.7 8.2 10.4 6.1 1.7 0.9

Sources: EU-LFS (spring data) and ECB calculations. Notes: For all tables on mismatch, the fi gures presented are those fulfi lling Eurostat's publications and reliability limits for the LFS data. The detailed education breakdown is only available for 2006. 1) 15 to 29 years old. 2) Differences in the total unemployment rate to table 4.1.a are due to missing data. 3) Starting 2004 for AT, BE, IE, PT, SE, UK. 4) Data for Ireland is 2005. 5) Average annual changes are presented in yellow. 6) EA calculated without IE.

105ECB


June 2008

ANNEX 3

Table 36 Unemployment rates and mismatch by level of educational attainment in the euro area and the euro area countries (15-29 year olds only) 1)

Country


Low Medium High


Range 2007

Change in range between highest and lowest rate (p.p)

1993-1995 1996-2001 2002-2007Belgium 23.9 14.2 7.5 13.9 16.4 8.4 2.8 0.2 0.0 -1.3 -0.2Germany 18.2 8.2 4.7 10.9 13.6 3.1 1.0 0.8 0.1 6.6 1.1Ireland 15.0 6.6 3.3 6.7 11.7 1.5 0.5 -12.2 7) -2.0 7) 3.47) 0.6 7)

Greece 14.3 17.3 19.2 17.0 4.9 -1.2 -0.4 -4.6 -0.8 1.4 0.2Spain 15.8 12.4 8.0 12.6 7.8 -4.2 -1.4 0.0 0.0 5.1 0.8France 4) 29.7 13.4 8.1 15.0 21.5 n.a. n.a 1.5 0.2 3.1 0.5Italy 15.5 12.1 14.4 13.4 3.4 3.0 1.0 -6.4 -1.1 0.9 0.1Luxembourg 15.2 7) n.a. n.a. 7.5 n.a. n.a. n.a n.a. n.a. n.a. n.a.Netherlands 5) 8.4 2.6 1.5 7) 4.5 6.9 7) n.a. n.a -5.3 -0.9 3.5 7) 0.6 7)

Austria 3) 11.5 5.4 n.a. 6.8 n.a. n.a. n.a 0.1 0.0 n.a. n.a.Portugal 13.1 12.2 14.0 13.0 1.8 -0.2 7) -0.1 7) -2.0 7) -0.3 7) 0.1 7) 0.0 7)

Slovenia 5) 13.1 7.3 5.8 7.6 7.2 n.a. n.a 5.1 7) 0.9 7) -8.4 7) -1.4 7)

Finland 3) 29.8 11.3 4.1 7) 15.7 25.7 7) n.a. n.a -4.8 -0.8 -4.4 7) -0.7 7)

Euro area 17.6 10.3 7.9 12.0 9.7 2.4 0.8 -1.6 -0.3 2.7 0.5Denmark 7.9 4.1 4.5 7) 5.8 3.8 -5.3 7) -1.7 7) 0.9 7) 0.2 7) 0.0 7) 0.0 7)

Sweden 3) 33.9 11.4 7.7 16.6 26.2 n.a. n.a -1.7 -0.3 12.3 2.1United

kingdom 21.6 9.2 3.6 10.4 18.0 -0.1 0.0 2.0 0.3 5.1 0.9

Sources: EU-LFS (spring data) and ECB calculations. Note: For all tables on mismatch, the fi gures presented are those fulfi lling Eurostat's publications and reliability limits for the LFS data. 1) The education levels refer to low: lower secondary education and less, medium: upper secondary education, high: tertiary education. In bold are the three best performers in terms of a low level and low range of unemployment rates.2) Education data start in 1992. 3) Data start in 1995. 4) Data start in 1993. 5) Data start in 1996. 6) Average annual changes are presented in the smaller font in yellow. 7) Based on fi gures smaller than the Eurostat reliability limit.

106ECB


June 2008

Table 37 Unemployment rates and mismatch by level of educational attainment in the euro area and the euro area countries (non-nationals, 25 to 64 year olds only) 1)


Country Low Medium High


Range2007

Change in range between highest andlowest rate (p.p) 7)

1996-2001 2002-2007

Belgium 25.2 14.6 7.9 16.0 17.3 -5.7 -0.9 6.0 1.0 Germany 21.1 13.4 11.3 16.2 9.7 1.3 0.2 4.8 0.8 Ireland 2) n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a.Greece 6.8 9.7 7.7 8.0 3.0 n.a. n.a. 0.7 0.1 Spain 13.9 10.4 10.7 12.0 3.6 -0.3 8) 0.0 8) -3.6 -0.6 France 20.6 15.2 n.a. 17.2 n.a. 0.6 0.1 n.a. n.a. Italy 3) 8.8 6.3 7.2 7.6 2.5 n.a. n.a. n.a. n.a. Luxembourg 4.6 n.a. 4.0 4.0 n.a. -1.3 -0.2 n.a. n.a. Netherlands 4) 11.3 5.4 6.1 6.8 5.9 n.a. n.a. n.a. n.a. Austria 16.4 8.7 n.a. 10.8 n.a. 1.0 0.2 n.a. n.a. Portugal 4) 18.3 n.a. n.a. 14.0 n.a. n.a. n.a. n.a. n.a. Slovenia 5) n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. Finland 29.2 14.3 10.4 17.9 18.8 n.a. n.a. n.a. n.a. Euro area 16.5 11.1 10.0 13.2 6.5 2.0 0.3 0.3 0.0 Denmark 14.5 n.a. n.a. 9.9 n.a. n.a. n.a. n.a. n.a. Sweden 6) 22.5 10.7 10.5 13.4 12.0 n.a. n.a. 5.5 0.9 United Kingdom 12.4 8.0 5.3 7.9 7.1 2.5 0.4 -2.1 -0.4

Sources: EU-LFS (spring data) and ECB calculations. Note: For all tables on mismatch, the fi gures presented are those fulfi lling Eurostat's publications and reliability limits for the LFS data. 1) Dataset starts 1995; the education levels refer to low: lower secondary education and less, medium: upper secondary education, high: tertiary education. 2) Only data for 1998-2004. 3) Only data for 2005-07. 4) Data start 1999. 5) Data start 2002. 6) Data start 1997. 7) Average annual changes are presented in yellow.8) Based on fi gures smaller than the Eurostat reliability limit.

107ECB


June 2008

ANNEX 3

Table 38 Unemployment rates and mismatch by level of educational attainment in the euro area and the euro area countries (non-nationals aged 15-29 years old) 1)

Country


Lower secondary education

and less

Upper secondary education

Tertiaryeducation

Weightedaverage ofsubgroups

Range2007

Change in range between highest and lowest rate (p.p) 7)

1996-2001 2002-2007

Belgium 36.2 23.7 8.8 24.5 27.4 2.1 0.3 12.4 2.1 Germany 24.9 12.8 9.9 17.8 15.0 0.6 0.1 7.7 1.3 Ireland 2) n.a. n.a. n.a. n.a. n.a. n.a. n.a n.a. n.a Greece 6.4 17.8 12.3 10.4 11.3 -3.1 -0.5 6.5 1.1 Spain 17.0 12.5 8.6 14.1 8.4 -12.6 -2.1 -0.1 0.0 France 32.7 19.2 15.9 24.4 16.7 3.1 0.5 -0.7 -0.1Italy 3) 10.3 9.3 12.2 10.1 2.8 n.a. n.a n.a. n.a Luxembourg 12.8 2.1 3.1 6.3 10.8 1.1 0.2 8.2 1.4 Netherlands 10.8 8.3 5.5 8.6 5.4 n.a. n.a -0.8 -0.1Austria 20.7 10.8 11.4 14.8 9.9 -7.2 -1.2 5.8 1.0 Portugal 4) 25.9 10.5 12.2 18.6 15.4 n.a. n.a -4.8 -0.8Slovenia 5) 0.0 0.0 n.a. 0.0 0.0 n.a. n.a n.a. n.a Finland 42.0 25.2 0.0 31.3 42.0 1.2 0.2 -2.4 -0.4Euro area 19.7 12.9 10.2 15.7 9.5 1.4 0.2 1.6 0.3 Denmark 19.0 10.5 11.9 13.2 8.5 47.9 8.0 -41.9 -7.0Sweden 6) 42.9 13.8 12.6 24.3 30.3 n.a. n.a 10.6 1.8 United Kingdom 15.4 10.5 7.2 10.4 8.3 1.9 0.3 -10.0 -1.7

Sources: EU-LFS (spring data) and ECB calculations. Note: For all tables on mismatch, the fi gures presented are those fulfi lling Eurostat's publications and reliability limits for the LFS data. 1) 15 to 29 years old; dataset starts 1995; the education levels refer to low: lower secondary education and less, medium: upper secondary education, high: tertiary education. 2) Only data for 1998-2004. 3) Only data for 2005-07. 4) Data start in 1999.5) Data start in 2002.6) Data start in 1997. 7) In yellow are average annual changes, a negative sign indicates that mismatch has decreased.

108ECB


June 2008

ANNEX 4

Box 12

CROSS-BORDER COMMUTING IN THE EURO AREA. CASE STUDY: LUXEMBOURG1

The number of outward-commuters from the euro area countries reached more than 2 million

persons in 2006 and has more than tripled over the past ten years.2 Commuters currently

represent 1.6% of euro area employment, with 87% of all commuters coming from only three

countries, namely Italy, Germany and France3. Outward-commuters are essentially men (83%)

aged between 25 and 54 (81%) with generally a low or medium level of education (respectively

43% and 41%). Over the last ten years, the share of cross-border commuters from euro area

countries with a “high” level of education (relative to those residing and working in the same

country) has remained relatively low.

Recent work analysing the pull and push factors on regional commuting fl ows in the European

Union 4 shows that commuting is well explained by the standard explanatory factors such as

the size of origin and destination regions and wage differentials. More specifi cally, high

unemployment and low wages in the home regions push workers towards commuting. For the

host country, cross-border labour supply results in a more effi cient allocation of labour across the

internal market. For the country of origin, commuters act as a buffer because outward commuting

reduces domestic unemployment in the case of asymmetric shocks. Moreover, commuting, as an

alternative to unemployment, allows workers to keep or even improve their skills, while reducing

unemployment benefi t-related expenditures. The home country also benefi ts from commuters’

incomes, which stimulate consumption.

Cross-border commuting in Luxembourg

Luxembourg does not have a continuous history of immigration, but has rather experienced three

distinct immigration waves: (i) Italians from the late nineteenth century to the 1950s, (ii) Portuguese

in the 1960s and 1970s - both characterised by a tendency towards permanent migration – and

(iii) since the 1980s, immigrants from a larger number of countries. The most recent estimates of

migration fl ows show Luxembourg as having the highest proportion of foreigner residents in its

population. In 2006, approximately 40% of the resident population were non-nationals. The more

recent phenomenon of a very rapidly growing number of commuters (the “frontaliers”) is closely

linked to opportunities in the labour market. In April 2007 about 39% of total employment and

68% of new jobs were occupied by cross-border workers.5 The expansion of the service industries,

notably fi nancial services and media-related companies during the last 20 years, has continued the

tendency for cross-border workers to be disproportionately over-represented in both low-skilled,

manual manufacturing jobs (55.5% of total employees in that sector) and construction (47.8%), in

high-skilled jobs in the fi nancial sector (48.4%) and in real estate, renting and business activities

1 Prepared by C. Olsommer.

2 Figures on outward commuting fl ows in the euro area are taken from Eurostat.

3 With respectively 1,373,284; 289,647 and 229,392 outward commuters in 2006.

4 J. Marvakov and T. Mathä (2006).

5 These high numbers are partly a refl ection of Luxembourg’s small size and geographical location in the middle of a large economic

area. The migration over such small distances would constitute only internal migration in Luxembourg’s larger neighbours. 51% of

cross-border immigrants are from France, 26% from Germany and 23% from Belgium.

109ECB


June 2008

ANNEX 4

(57.2%). Average educational attainment differs considerably among the various nationalities of

foreign residents and “frontaliers” working in Luxembourg.

The most recent studies about the determinants of cross-border migration in Luxembourg

emphasise the neo-classical theory that individuals’ migration decisions are determined partly

by the expected income from work. Average monthly wages and the minimum wage levels in

Luxembourg are higher than in the neighbouring countries.6 This wage differential between the

expected wage in Luxembourg and in the residence country might be especially important for

commuters, since they are likely to spend most of their wage in their country of residence.7 Other

factors infl uencing the cross-border commuting decision include the fi nancial and non-fi nancial

costs of such temporary migration. Lastly, a high probability of fi nding a job,8 combined with

access to social security benefi ts (in particular child benefi ts and relatively generous parental

leave) increase the advantages relative to costs of temporary migration to Luxembourg.

These migrant fl ows have an impact on labour supply in both Luxembourg and its neighbouring

countries. Luxembourg experiences a shortage of skilled labour, with more than a half of its

resident unemployed being low skilled.9 Temporary immigration helps to reduce labour shortages,

particularly in the fi nancial sector by supplying specialised and high-skilled labour. Commuters

therefore complement labour supply from the resident labour force and facilitate domestic factor

utilisation. Neighbouring countries also benefi t from this temporary migration, since jobs seekers,

by fi nding a job in Luxembourg, reduce the unemployment rate in their country of residence.10

Luxembourg’s experience, despite some traffi c problems, provides a successful example of

cross-border labour mobility helping to complement the domestic labour force.

6 OECD annual data show the monthly minimum wage (in PPS) in 2005 to be: €1,417 for unskilled workers and €1,700 for skilled

workers in Luxembourg, €1,218 in France and €1,234 in Belgium.

7 Although, of course, this wage differential may be more prominent in certain sectors and thus more relevant for particular groups of

potential commuters.

8 Job creation was vigorous in Luxembourg over the past twenty years, reaching on average 3.6% per year. The cross-border workers

largely benefi ted from this, as they occupied 7/10ths of the new jobs on average since 1986.

9 Luxembourg’s Beveridge curve shifted out during the 1990s and has done so again since 2005, indicating an increasing degree of

mismatch between supply and demand. This conclusion is also supported by alternative measures of structural change, such as the

Lilien indicator of inter-sectoral structural change and the rate of unsatisfi ed sectoral labour demand over unsatisfi ed sectoral labour

supply. See also Banque centrale du Luxembourg (2004).

10 All the more since the unemployment rates in neighbouring states are higher than in Luxembourg (4.5% in 2006). Lorraine (France):

11.0%, Saarland (Germany): 9.7%, Rhineland-Palatinate (Germany): 8.2% and Wallonia (Belgium): 18.8%.

110ECB


June 2008

REFERENCES

Aaronson, S., B. Fallick, A. Figura, J. Pingle and W. Wascher (2006), “The Recent Decline in the

Labour Force Participation Rate and Its Implications for Potential Labour Supply”, Brookings

Papers on Economic Activity, March.

Acemoglu, D. and J.S. Pischke (1998), “Why do Firms Train? Theory and Evidence”, Quarterly

Journal of Economics, Vol. 113:1, pp. 79-119.

Acemoglu, D. and J.S. Pischke (1999), “Beyond Becker: Training in Imperfect Labour Markets”,

Economic Journal, Vol. 109:453, pp. F112-F142.

Acemoglu, D., D.H. Autor and D. Lyle (2002), “Women, War and Wages: The Effect of Female

Labor Supply on the Wage Structure at Mid-century”, NBER Working Papers, No 9013.

Acemoglu, D. (2003), “Cross-Country Inequality Trends”, Economic Journal, February 2003, Vol 113,

pp. F121-149.

Aghion, P, M. Dewatripont, C. Hoxby, A. Mas-Colell and A. Sapir (2007), “Why reform Europe’s

universities?”, Bruegel policy brief, September 2007/04.

Amable, B., L. Demmou and D. Gatti (2007), “Employment Performance and Institutions:

New Answers to an Old Question”, IZA Discussion Paper, No 2731.

Ammermüller, A. and A.M. Weber (2005), “Educational Attainment and Returns to Education

in Germany”, ZEW-Discussion Paper, No 05-17.

Amuedo-Dorantes, C. and S. de la Rica (2007), “Labour Market Assimilation of Recent Immigrants

in Spain”, forthcoming in the British Journal of Industrial Relations.

Antunes, A. and T. Cavalcanti (2007), “Start Up Costs, Limited Enforcement, and the Hidden

Economy”, European Economic Review 51(1), pp. 203-24.

Ashenfelter, O., C. Harmon and H. Oosterbeek (1999), “A Review of Estimates of the

Schooling/Earnings Relationship, with Tests for Publication Bias”, Labour Economics 6(1),

pp. 453-470.

Baker, M. and D. Benjamin (1994), “The Performance of Immigrants in the Canadian Labour

Market”, Journal of Labour Economics, Vol. 12, pp.369-405.

Bank of Italy (2006), Annual Report.

Banque centrale du Luxembourg (2004) “(Le) Chômage structurel (encadré)”, in Bulletin de

la Banque centrale du Luxembourg, n° 4/2004, pp.36-40.

Barrett, A., A. Bergin and D. Duffy (2006), “The Labour Market Characteristics and Labour Market

Impacts of Immigrants in Ireland”, IZA Discussion Paper, No 1553 and Economic and Social

Review Vol.37, No 1.

111ECB


June 2008

REFERENCES

Barrett, A. and Y. McCarthy (2007), “Immigrants in a Booming Economy: Analysing their Earnings

and Welfare Dependence”, Labour: Review of Labour Economics and Industrial Relations,

Vol. 21 (4-5), pp. 789-808.

Barro R. (1991), “Economic growth in a cross section of countries”, Quarterly Journal of Economics,

May, 106(2), pp. 407-43.

Barro R. and X. Sala-I-Martin (1995), “Capital mobility in neoclassical models of growth”,

American Economic Review, 85, pp. 103-115.

Barro R. (2001), “Human capital and growth”, American Economic Review, May, 91(2), pp.12-17.

Barro R. and X. Sala-I-Martin (2004), “Economic Growth”, MIT Press, Cambridge.

Bassanini, A. and R. Duval (2006), “Employment Patterns in OECD Countries: Reassessing the

Role of Policies and Institutions”, OECD Social, Employment and Migration Working Papers

No 35, Paris.

Becker, G. (1962), “Investment in human capital: a theoretical analysis”, Journal of Political

Economy, Vol. 70:5, pp. 9-49.

Benhabib J. and M.M. Spiegel (1994), “The role of human capital in economic development:

Evidence from aggregate cross-country data”, Journal of Monetary Economics, October, 34(2),

pp. 143-74.

Berger, M.C. (1985), “The Effect of Cohort Size on Earnings Growth: A Reexamination of the

Evidence”, The Journal of Political Economy, Vol. 93, No 3, pp. 561-573.

Bertola, G., F. Blau and L. Kahn (2002), “Labor Market Institutions and Demographic Employment

Patterns”, NBER Working Paper, No 9043.

Blanchard, O. and J. Wolfers (2000), “The Role of Shocks and Institutions in the Rise of European

Unemployment: The Aggregate Evidence”, The Economic Journal, Vol. 110, March.

Blanchard and Gali (2007), “Real wage rigidities and the New Keynesian Model”, Journal

of Money, Credit and Banking, Vol. 39:2, pp.35-66

Blöndal, S. and S. Scarpetta (1999), “The Retirement Decision in OECD Countries”, OECD

Working Paper, No 202.

Blundell, R. and T. Macurdy (1999), “Labor Supply: A Review of Alternative Approaches”,

in Ashenfelter, O. and Card. D. (eds), Handbook of Labor Economics 3, pp. 1559-1695.

Borjas, G. (2000), “Foreign-Born Teaching Assistants and the Academic Performance

of Undergraduates,” American Economic Review, Vol. 90(2), pp. 355-359.

Borjas, G. (2001), “Heaven’s Door”, Princeton University Press.

112ECB


June 2008

Bosworth, B.P. and S.M. Collins (2003), “The empirics of growth: An update”, Brookings Papers

on Economic Activity 2003, No 2, pp. 113-206.

Bradbury K. (2005), “Additional slack in the economy: the poor recovery in labor force participation

during this business cycle”, Federal Reserve Bank of Boston, No 05-2.

Brandolini, A. and P. Cipolline (2001), “Multifactor Productivity and Labour Quality in Italy,

1981-2000”, Temi di discussione, Banca d’Italia.

Brandolini, A, P. Cipollone and E. Viviano (2006), “Does ILO Defi nition Capture all

Unemployment?”, Journal of the European Economic Association, 4(1), pp. 153-179.

Budría, S and P.T. Pereira (2005), “Educational Qualifi cations and Wage Inequality: Evidence for

Europe,” IZA Discussion Paper, No 1763.

Burda, M. D. Hamermesh and P. Weil (2006), “The Distribution of Total Work in the EU and US”,

IZA Discussion Paper, No 2270.

Caliendo, M., A. Kritikos, V. Steiner and F. Wießner (2007), “Existenzgründungen: Unterm Strich

ein Erfolg”, IAB Kurzbericht 10, IAB.

Callan, T., A. van Soest and J. Walsh (2003), “Tax cuts, tax reform and labour supply responses”,

Study for the Social Economic Research Institute, Dublin.

Callan, T., A. van Soest and J.R. Walsh (2007), “Tax Structure and Female Labour Market

Participation: Evidence from Ireland,” IZA Discussion Paper, No 3090.

Card, D. and R. Freeman (2004), “What Have Two Decades of British Economic Reform

Delivered?”, in Card, D., R. Blundell and R. Freeman (editors): Seeking a Premier Economy:

The Economic Effects of British Economic Reforms, 1980-2000, NBER.

Carnoy, M., M. Castells and C. Brenner (1997), “Labour Markets and Employment Practices in the

Age of Flexibility”, International Labour Review, Vol.136, Spring.

Carone, G., and A. Salomäki (2001), “Reforms in Tax-Benefi t Systems in order to Increase

Employment Incentives in the EU”, DG-ECFIN Economic Paper 160.

Carrasco, R., J.F Jimeno and A.C. Ortega (2008), “The effects of immigration on the labour market

performance of native-born workers: some evidence for Spain”, forthcoming in Journal of

Population Economics, Vol.3/2008.

Centeno, M. and P. Fernandes (2004), “Labour Market Heterogeneity: Distinguishing between

Unemployment and Inactivity”, Banco de Portugal, Economic Bulletin.

Centeno, M. and A. Novo (2007), “Identifying Unemployment Insurance Income Effects with a

Quasi-Natural Experiment”, Banco de Portugal Working Paper, No 10.

113ECB


June 2008

REFERENCES

Chiswick, B. R. and P.W. Miller (1995), “The Endogeneity between Language and Earnings:

International Analyses,” Journal of Labor Economics, Vol. 13(2), pp. 246-288.

Chiswick, B. R. and P.W Miller (2005), “Why is the Payoff to schooling smaller for immigrants?”,

IZA Discussion Paper, No 1731.

Christoffel, K.P., K. Kuester and T. Linzert (2006), “Identifying the role of labor markets for monetary

policy in an estimated DSGE model”, Working Paper Series, No 635, European Central Bank.

Ciccone, A., and E. Papaioannou (2005), “Human Capital, the Structure of Production, and

Growth”, Barcelona, Universitat Pompeu Fabra.

Clark K.B. et al. (1979), “Labor Market Dynamics and Unemployment: A Reconsideration”,

Brookings Papers on Economic Activity, No 1, pp.13-72.

CNAF (2005), “La PAJE après 18 mois de montée en charge”, L’e-ssentiel, n°42.

Copenhagen Economics (2007) “Study on reduced VAT applied to goods and services in the

Member States of the European Union”, Kopenhagen. www.ec.europa.eu/taxation_customs/

resources/documents/taxation/vat/how_vat_works/rates/study_reduced_VAT.pdf

Coulombe, S., J.F. Tremblay and S. Marchand (2004), “Literacy scores, human capital and growth

across fourteen OECD countries”, Statistics Canada, Ottawa.

Coulombe, S. and J.F. Tremblay (2006), “Literacy and Growth”, Topics in Macroeconomics 6, No 2.

Daly, Kevin (2007), “Gender inequality, growth and global ageing”, Goldman Sachs Global

economics paper, No. 154.

Deutsche Bundesbank (2006), “The evolution of labour market related-expenditure”, Monthly

Bulletin, September.

Duval, R. (2003), “The retirement effects of old-age pension and early retirement schemes in OECD

countries”, OECD Working Paper, No 304.

Duval, R. (2006), “Pension Systems, Social Transfer Programmes and the Retirement Decision

in OECD Countries”, in Messina, J., C. Michelacci, J. Turunen G. and Zoega (eds.)

Labour Market Adjustments in Europe, Edward Elgar, Cheltenham.

Duval R., J. Elmeskov and L. Vogel (2007), “Structural Policies and Economic Resilience to

Shocks”, OECD Economics Department Working Papers, No 567, OECD.

Easterlin, R. (1980), “Birth and Fortune: The Effects of Generation Size on Personal Welfare”,

New York: Basic Books.

Eichhorst, W. and K.F. Zimmermann (2007), “How many (and Which) Measures of Active Labor

Market Policy Do We Still Need?”, IZA Discussion Paper, No 2605.

114ECB


June 2008

Elmeskov J. and K. Pichelmann (1993), “Interpreting unemployment: the role of labour-force

participation”, OECD Economic Studies, No 21.

European Central Bank (2002), “Labour Market Mismatches in Euro Area Countries”, March,

Frankfurt.

European Central Bank (2006), “Developments in the structural features of euro area labour markets

over the last decade”, Monthly Bulletin, September.

European Commission (2002a), “Social Situation Report 2002”, Brussels.

European Commission (2002b), “The Gender Pay Gap – Origins and policy responses:

A comparative review of 30 European countries”, Paper by a group of experts on Gender, Social

Inclusion and Employment, July 2006.

European Commission (2006), “The impact of ageing on public expenditure: Projections for the

EU25 Member States on pensions, health-care, long-term care, education and unemployment

transfers (2004-2050)”, European Economy, Special Report No 1.

Employment Committee (2007), “Employment Committee Working Group Report on Active

Ageing”, http://ec.europa.eu/employment_social/employment_strategy/emco/activeageing_

workinggroup07_en.pdf.

Eurostat (1996), “The European Union Labour Force Survey: Methods and Defi nitions”, Offi ce for

Offi cial Publications of the European Communities, Luxembourg.

Eurostat (2007), “Reconciliation between work and family life”.

Fallick, B. and J. Pingle (2007), “A Cohort Based Model for Labour Force Participation”, Finance

and Economics Discussion Series, Board of Governors of the Federal Reserve System.

Fernández, C. and A.C. Ortega (2006), “Labour Market Assimilation of Immigrants in Spain:

Employment at the Expense of Bad Job-Matches?”, IESE Business School working paper, No 644.

Finegold, D and D. Soskice (1988), “The failure of training in Britain: analysis and prescription”,

Oxford Review of Economic Policy, Vol.4:3, pp.21-53.

Fiori, G., G. Nicoletti, S. Scarpetta, and F. Schiantarella (2007), “Employment Outcomes and

the Interaction Between Product and Labor Market Deregulation: Are They Substitutes or

Complements?”, IZA Discussion Paper, No 2770.

Freeman, R.B. and R. Schettkat (2005), “Marketization of Household Production and the

EU-US Gap in Work”, Economic Policy, pages 5-50.Genre, V., R. Gomez-Salvador and A. Lamo

(2005b), “The Determinants of Labour Force Participation in the EU” in R. Gómez-Salvador,

A. Lamo, B. Petrongolo, M. Ward, and E. Wasmer (eds.) Labour Supply and Incentives to Work

in Europe, Edward Elgar.

115ECB


June 2008

REFERENCES

Genre, V. and R. Gomez-Salvador (2002), “Labour force developments in the euro area since the

1980s” ECB Occasional Paper, No 4.

Genre, V., R. Gomez-Salvador and A. Lamo (2007), “European Women. Why do(n’t) they work?”,

Applied Economics, forthcoming.

Gomez-Salvador, R., A. Musso, M. Stocker and J. Turunen (2006), “Labour productivity

developments in the euro area”, ECB Occasional Paper, No 53.

González L. (2005), “Nonparametric Bounds on the Returns to Language Skills”, Journal of Applied

Econometrics, Vol 20(6), pp. 771-795.

Gronau, R. and Y. Weiss (1981), “Expected Interruptions in Labour Force Participation and

Sex-Related Differences in Earnings Growth”, Review of Economic Studies, Vol. 48(4),

pp. 607-19.

Gustafsson, S. and F.P. Stafford (1992), “Child care subsidies and Labour supply in Sweden”,

Journal of Human Resources, Vol. 27, no.1, pp. 204-230.

Hanushek E. and D. Kimko (2000), “Schooling, labour-force quality, and the growth of nations”,

American Economic Review, December, 90(5), pp. 1184-1208.

Hanushek, E and L. Wößmann (2007), “The role of education quality for economic growth”, World

Bank Policy Research, Working Paper, No 4122, Washington, World Bank.

Inklaar, R., M. O’Mahony and M. Timmer (2005), “ICT and Europe’s Productivity Performance:

Industry Level Growth Account Comparisons with the United States”, Review of Income and

Wealth 51(4), pp. 505-536.

International labour review (1997), International Labour Organization.

Jacobi, L. and J. Kluve (2007), “Before and after the Hartz reforms: The performance of active

labour market policy in Germany”, Zeitschrift für Arbeitsmarktforschung, 40, 1, pp. 45-64.

Jaumotte, F. (2003), “Female Labour Force Participation: Past Trends and Main Determinants in

OECD Countries”, OECD Economics Department Working Papers No 376.

Jimeno, J.F. and D. Rodriguez-Palenzuela (2002), “Youth unemployment in the OECD:

demographic shifts; labour market institutions, and macroeconomic shock”, Working Paper,

No 155, European Central Bank.

Jorgenson, D. (2005), “Accounting for Growth in the Information Age”, in P. Aghion and

S. Durlauf, eds., Handbook of Growth Economics, Vol 1A, pp.743-815.

Jones, S.R.G. and W.C. Riddell (1999), “The Measurement of Unemployment: An Empirical

Approach”, Econometrica, 67, pp. 147-162.

Kailis, E. and S. Pilos (2005), “Lifelong learning in Europe”, Eurostat Statistics in Focus, 8/2005.

116ECB


June 2008

Kanellopoulos, K., K. Mavromaras and T. Mitrakos (2004), “Education and the labour market”,

Study No 50, KEPE: Athens (in Greek).

Kate C., H. Bachnan and F. J. Morrison (2001), “Schooling and Cognitive Development,” in

Sternberg, R.J. and E.L. Grigorenko (eds.) Environmental Effects on Cognitive Abilities,

Mahwah, New Jersey: Lawrence Erlbaum Associates, Publishers, Chapter 10, pp. 287 - 335.

Kettner, A. and M. Rebien (2007), “Hartz-IV-Reform: Impulse für den Arbeitsmarkt“,

IAB Kurzbericht 19, IAB.

Killingsworth, M. and J. Heckman (1986), “Female Labor Supply: A Survey”, in: Ashenfelter,

O. and Layard, R. (eds), Handbook of Labor Economics (1), North-Holland, Amsterdam,

pp. 103-204.

Korenman, S. and D. Neumark (1997), “Cohort Crowding and Youth Labor Markets: A

Cross-National Analysis”, NBER Working Paper, No 6031.

Kossoudji, S.A. and D. Cobb Clark (2002), “Coming out of the Shadows: Learning about Legal

Status and Wages from the Legalized Population”, Journal of Labor Economics, Vol. 20(3),

pp. 598-628.

Lauer, C and V. Steiner (2000), “Private Erträge von Bildungsinvestitionen in Deutschland”,

Beihefte der Konjunkturpolitik, Applied Economics Quarterly, 51, pp. 71-101.

Layard R., S. Nickell and R. Jackman (2005), ”Unemployment”, Chapter 8, Oxford University

Press.

Leiner-Killinger N., C. Madaschi and M. Ward-Warmedinger (2005), “Trends and patterns in

working time across euro area countries 1970-2004: causes and consequences”, ECB Occasional

Paper No 41, December 2005.

Levenson, A. and C. Zoghi (2007), “The Strength of Occupation Indicators as a Proxy for Skill”

Working Papers 404, U.S. Bureau of Labor Statistics.

Lilien, D. and R.E. Hall (1986), “Cyclical Fluctuations in the Labor Market”, in 0. C.Ashenfelter

and R. Layard, eds., Handbook of Labor Economics, Vol. 2, Amsterdam: North-Holland.

Lipsey, R.G. (1960), “The Relation between Unemployment and the Rate of Change of Money

Wage Rates in the United Kingdom, 1862-1957: A Further Analysis”, Economica, Vol. 27,

No 105. (Feb., 1960), pp. 1-31.

Lopez-Garcia, P. (2003) “Labour Market Performance and Start-up Costs: OECD evidence”,

mineo.

Maddaloni, A., A. Musso, P. Rother, M. Ward-Warmedinger and T. Westermann (2006),

“Macroeconomic Implications of Demographic Developments in the Euro Area”, European

Central Bank Occasional Paper, No 51.

117ECB


June 2008

REFERENCES

Magoula, T. and G. Psacharopoulos (1999), “Schooling and monetary rewards in Greece: an

over-education false alarm?”, Applied Economics 31: 12, pp.1589-97.

Martins, P., A. Novo and P. Portugal (2007), “Increasing the Legal Retirement Age: The Impact on

Wages, Hours, Worker Flows and Firm Performance”, Banco de Portugal, mimeo.

Marvakov, J. and T.Y. Mathä (2007), “An analysis of regional commuting fl ows in the European

Union”, Cahier d’Etudes n°28, Banque Centrale du Luxembourg.

Melka, J. and L. Nayman (2004), “Labour Quality and Skill Biased Technological Change in

France”, CEPII, unpublished.

Micheaux, S and O. Monso (2007), “Faire garder ses enfants pendant son temps de travail”, Insee

Working Paper, No 1132, April.

Mincer J. and S. Polachek (1974), “Family investments in human capital: earnings of women”,

Journal of Political Economy, Vol. 82 (supplement), pp. S76-S108.

Montanaro P. (2007), “Quality of education and human capital: An overview”, Bank of Italy

discussion paper.

National Research Council, Board on Children and Families, Commission on Behavioral and Social

Sciences and Education (1995), “Immigrant Children and Their Families: Issues for Research

and Policy,” The Future of Children, No 5 (2 Summer/Fall), pp.72 - 89.

Nickell S. and R. Layard (1999), “Labor Market Institutions and Economic Performance”,

in: Ashenfelter, O. and D. Card (eds.), Handbook of Labor Economics (3) North-Holland,

Amsterdam, Chapter 42, pp. 3029-3084.

Nicoletti, G., A. Bassanini, E. Ernst, S. Jean, P. Santiago and P. Swa (2001), “Product and

Labour Markets Interactions in OECD Countries”, OECD Economics Department Working

Paper No 312.

Nicoletti, G. and S. Scarpetta (2005), “Product Market Reforms and Employment in OECD

Countries”, OECD Economics Department Working Paper No 472.

OECD (1998), “Employment Outlook”, Paris.

OECD (2001a), “Employment Outlook”, Paris.

OECD (2001b), OECD Productivity Manual: A Guide to the Measurement of Industry-Level and

Aggregate Productivity Growth, Paris.

OECD (2006a), “Education at a Glance”, Paris.

OECD (2006b), “Employment Outlook”, Paris.

OECD (2006c), “International Migration Outlook”, Sopemi Edition, Paris.

118ECB


June 2008

OECD (2006d), “Taxing wages 2005-2006”, Paris.

OECD (2007a), “Babies and Bosses. Reconciling Work and Family Life. A Synthesis of Findings

for OECD Countries”, Paris.

OECD (2007b), “Benefi ts and Wages”, Paris.

OECD (2007c), “Economic Surveys: Austria”, Paris.

OECD (2007d), “Education at a glance”, Paris.

OECD (2007e), “Employment Outlook”, Paris.

OECD (2007f), “International Migration Outlook”, Paris.

OECD (2007g), “Pensions at a glance”, Paris.

OECD (2007h), “Taxing wages 2005-2006”, Paris.

Papapetrou, E. (2007), “Education, labour market and wage differentials in Greece”, Bank of

Greece Economic Bulletin, No 28, pp. 51-73.

Pencavel, J. (1986), “Labor Supply of Men: a Survey”, in: Ashenfelter, O. and R. Layard (eds.),

Handbook of Labor Economics (1), North-Holland, Amsterdam, pp. 3-102.

Petrongolo B. and C. Pissarides (2001), “Looking into the black box: A Survey of the Matching

Function”, 39, pp. 390-431.

Pissarides, C., P. Garibaldi, C. Olivetti, B. Petrolongo and E. Wasmer (2003), “Women in the

Labour Force: How Well is Europe Doing?”, in Boeri, T., C. Pissarides and D. Del Boca (eds.)

Women at Work: an Economic Perspective, Oxford University Press, Oxford.

Rivera-Batiz, F.L. (1999), “Undocumented workers in the labor market: An analysis of the earnings

of legal and illegal Mexican immigrants in the United States”. Journal of Population Economics,

Vol. 12(1), pp. 91-116.

Rogerson, R. (2003), “Structural transformation and the deterioration of European Labour Market

Outcomes”, mimeo, Arizona State University.

Rogerson, R. (2007), “Taxation and market work: is Scandinavia an outlier?” NBER Working

Paper, No 12890.

Romans, F. (2007), “The transition of women and men from work to retirement”, Eurostat Statistics

in Focus 97/2007.

Sala-i-Martin, X. (1997), “I just Ran Four Million Regressions”, Economics Working Papers 201,

Department of Economics and Business, Universitat Pompeu Fabra.

119ECB


June 2008

REFERENCES

Schneider. F. (2006), “Shadow economies and corruption all over the world: What do we really

know?”, IZA Discussion Paper, No 2315, Bonn.

Schneider, H., K. Brenke, B. Jesske, L. Kaiser, U. Rinne, M. Schneider, J. Steinwede and

A. Uhlendorff (2007), “Evaluation der Maßnahmen zur Umsetzung der Vorschläge der

Hartz-Kommission: Bericht 2006”, IZA Research Report No 10.

Schwerdt, G. and J. Turunen (2007), “Growth in Euro Area Labour Quality”, Review of Income

and Wealth 53(4), pp. 716-734.

Shimer, R. (1998), “Why is the US unemployment rate so much lower?”, NBER Macroeconomics

Annual, pp. 13-73.

Silva, A. (2005), “Taxes and Labor Supply: Portugal, Europe and the United States”, Banco de

Portugal, mimeo.

Social Security Administration (2006), “Social Security Programs Throughout the World: Europe,

2002”, Offi ce of Research, Evaluation and Statistics, Washington.

Steiner, V. and C. Lauer (2000), “Private Erträge von Bildungsinvestitionen in Deutschland,

Beihefte der Konjunkturpolitik”, Applied Economics Quarterly 51, pp. 71-101.

Steiner V. and K. Wrohlich (2006), “Introducing Family Tax-Splitting in Germany: How Would it

Affect the Income Distribution and Work Incentives?”, IZA Discussion Paper, No 2245.

Stevens, M. (1996), “Transferable training and poaching externalities”, In Booth, A.L. and D.J. Snower

(eds.) Acquiring skills, Cambridge University Press.

Stevens, M. (2001), “Should fi rms be required to pay for vocational training”, Economic Journal,

Vol.111:743, pp. 485-505.

Stock, J.H. and D.A. Wise (1990), “Pension, the option value of work and retirement”, Econometrica

Vol. 58:5, pp. 1151-1180.

Stoiber, S.A. (1990), “Family Leave Entitlements in Europe: Lessons for the United States”,

Compensation and Benefi ts Management, VI.

Tam, M. (1997), “Part-time employment: A bridge or a trap?”, Aldershot, Avebury.

The Economic Bureau of the President (2006), “Immigration and the Spanish economy” http://

www.la-moncloa.es/NR/rdonlyres/88165A8C-E40D-4D7E-9468-FB0C45516947/81221/

InmigrationandSpanishEconomy.pdf

Vamvakidis, A (2008) “Regional Wage differentiation and Wage Bargaining Systems in the EU”,

IMF Working Paper, No 08/43.

Viitanen, T.K. (2005), “Cost of Childcare and Female Employment in the UK”, Labour, Vol. 19,

No S1, pp. 149-170.

120ECB


June 2008

Wößmann, L. (2007), “International Evidence on School Competition, Autonomy and

Accountability: A Review”, Peabody Journal of Education, Vol. 82, No 2-3, pp.473-497.

Weichselbaumer, D. and R. Winter-Ebmer (2005), “A Meta-Analysis of the International Gender

Wage Gap”, Journal of Economic Surveys, 9 (3), pp. 479-511.

Welch, F. (1979), “Effects of Cohort Size on Earnings: The Baby Boom Babies’ Financial Bust”,

Journal of Political Economy, Vol. 87, No 5.

121ECB


June 2008

EUROPEAN

CENTRAL BANK

OCCAS IONAL

PAPER SERIES

EUROPEAN CENTRAL BANK

OCCASIONAL PAPER SERIES SINCE 2007

55 “Globalisation and euro area trade: Interactions and challenges” by U. Baumann and F. di Mauro,

February 2007.

56 “Assessing fi scal soundness: Theory and practice” by N. Giammarioli, C. Nickel, P. Rother,

J.-P. Vidal, March 2007.

57 “Understanding price developments and consumer price indices in south-eastern Europe”

by S. Herrmann and E. K. Polgar, March 2007.

58 “Long-Term Growth Prospects for the Russian Economy” by R. Beck, A. Kamps and

E. Mileva, March 2007.

59 “The ECB Survey of Professional Forecasters (SPF) a review after eight years’ experience”,

by C. Bowles, R. Friz, V. Genre, G. Kenny, A. Meyler and T. Rautanen, April 2007.

60 “Commodity price fl uctuations and their impact on monetary and fi scal policies in Western and

Central Africa” by U. Böwer, A. Geis and A. Winkler, April 2007.

61 “Determinants of growth in the central and eastern European EU Member States – A production

function approach” by O. Arratibel, F. Heinz, R. Martin, M. Przybyla, L. Rawdanowicz,

R. Serafi ni and T. Zumer, April 2007.

62 “Infl ation-linked bonds from a Central Bank perspective” by J. A. Garcia and A. van Rixtel,

June 2007.

63 “Corporate fi nance in the euro area – including background material”, Task Force of the

Monetary Policy Committee of the European System of Central Banks, June 2007.

64 “The use of portfolio credit risk models in central banks”, Task Force of the Market Operations

Committee of the European System of Central Banks, July 2007.

65 “The performance of credit rating systems in the assessment of collateral used in Eurosystem

monetary policy operations” by F. Coppens, F. González and G. Winkler, July 2007.

66 “Structural reforms in EMU and the role of monetary policy – a survey of the literature”

by N. Leiner-Killinger, V. López Pérez, R. Stiegert and G. Vitale, July 2007.

67 “Towards harmonised balance of payments and international investment position statistics – the

experience of the European compilers” by J.-M. Israël and C. Sánchez Muñoz, July 2007.

68 “The securities custody industry” by D. Chan, F. Fontan, S. Rosati and D. Russo, August 2007.

69 “Fiscal policy in Mediterranean countries – Developments, structures and implications for

monetary policy” by M. Sturm and F. Gurtner, August 2007.

122ECB


June 2008

70 “The search for Columbus’ egg: Finding a new formula to determine quotas at the IMF”

by M. Skala, C. Thimann and R. Wölfi nger, August 2007.

71 “The economic impact of the Single Euro Payments Area” by H. Schmiedel, August 2007.

72 “The role of fi nancial markets and innovation in productivity and growth in Europe”

by P. Hartmann, F. Heider, E. Papaioannou and M. Lo Duca, September 2007.

73 “Reserve accumulation: objective or by-product?” by J. O. de Beaufort Wijnholds and

L. Søndergaard, September 2007.

74 “Analysis of revisions to general economic statistics” by H. C. Dieden and A. Kanutin,

October 2007.

75 “The role of other fi nancial intermediaries in monetary and credit developments in the euro area”

edited by P. Moutot and coordinated by D. Gerdesmeier, A. Lojschová and J. von Landesberger,

October 2007.

76 “Prudential and oversight requirements for securities settlement a comparison of cpss-iosco”

by D. Russo, G. Caviglia, C. Papathanassiou and S. Rosati, November 2007.

77 “Oil market structure, network effects and the choice of currency for oil invoicing” by E. Mileva

and N. Siegfried, November 2007.

78 “A framework for assessing global imbalances” by T. Bracke, M. Bussière, M. Fidora and

R. Straub, January 2008.

79 “The working of the eurosystem: monetary policy preparations and decision-making – selected

issues” by P. Moutot, A. Jung and F. P. Mongelli, January 2008.

80 “China’s and India’s roles in global trade and fi nance: twin titans for the new millennium?”

by M. Bussière and A. Mehl, January 2008.

81 “Measuring Financial Integration in New EU Member States” by M. Baltzer, L. Cappiello,

R.A. De Santis, and S. Manganelli, January 2008.

82 “The Sustainability of China’s Exchange Rate Policy and Capital Account Liberalisation”

by L. Cappiello and G. Ferrucci, February 2008.

83 “The predictability of monetary policy” by T. Blattner, M. Catenaro, M. Ehrmann, R. Strauch

and J. Turunen, March 2008.

84 “Short-term forecasting of GDP using large monthly datasets: a pseudo real-time forecast

evaluation exercise” by G. Rünstler, K. Barhoumi, R. Cristadoro, A. Den Reijer, A. Jakaitiene,

P. Jelonek, A. Rua, K. Ruth, S. Benk and C. Van Nieuwenhuyze, May 2008.

85 “Benchmarking the Lisbon Strategy” by D. Ioannou, M. Ferdinandusse, M. Lo Duca, and

W. Coussens, June 2008.

123ECB


June 2008

EUROPEAN

CENTRAL BANK

OCCAS IONAL

PAPER SERIES

86 “Real convergence and the determinants of growth in EU candidate and potential candidate

countries: a panel data approach” by M. M. Borys, É. K. Polgár and A. Zlate, June 2008.

87 “Labour supply and employment in the euro area countries: developments and challenges”,

by a Task Force of the Monetary Policy Committee of the European System Of Central Banks,

June 2008.

Task force of the Monetary Policy Committee of the European System of Central Banks

Occas iOnal PaPer ser i e snO 87 / J une 2008

nO

87

Ju

ne

2008

Occ

asiO

nal

PaP

er s

erie

s

laBOur suPPlY anD

eMPlOYMenT in THe

eurO area cOunTries

DeVelOPMenTs

anD cHallenGes

Labour supply and employment in the euro area countries - developments and challenges

Documents

Transcript of Labour supply and employment in the euro area countries - developments and challenges